WO2022246767A1

WO2022246767A1 - Method and device for determining steering intention of target vehicle

Info

Publication number: WO2022246767A1
Application number: PCT/CN2021/096541
Authority: WO
Inventors: 文韬
Original assignee: 华为技术有限公司
Priority date: 2021-05-27
Filing date: 2021-05-27
Publication date: 2022-12-01
Also published as: CN113226885A

Abstract

A method and device for determining a steering intention of a target vehicle. According to the method, pose information of a target vehicle is determined according to visual information of the target vehicle, such that a steering intention of the target vehicle is further determined according to the pose information of the target vehicle. Hence, by acquiring pose information of a target vehicle according to image information, the pose information of the target vehicle can represent a steering intention of the target vehicle more accurately, and the steering intention of the target vehicle determined using the method can be more accurate, thereby facilitating improving the accuracy of predicting the steering intention of the target vehicle, better assisting automatic driving, and then improving the safety of driving an autonomous vehicle.

Description

一种目标车辆转向意图的确定方法及装置A method and device for determining the steering intention of a target vehicle

技术领域technical field

本申请涉及自动驾驶技术领域，特别涉及一种目标车辆转向意图的确定方法及装置。The present application relates to the technical field of automatic driving, and in particular to a method and device for determining a steering intention of a target vehicle.

背景技术Background technique

近年来自动驾驶技术和高级驾驶辅助***进入飞速发展时期，自动驾驶车辆在道路上的行驶安全性也因此成为了关注的重点。自动驾驶车辆在道路上的安全行驶要求自动驾驶车辆能准确感知、理解和预测各类交通参与者的行驶轨迹或转向意图，以做出安全合理的避障决策。预测目标车辆的转向意图的精准度，关系着自动驾驶车辆在道路上的行驶安全性。如何提高预测目标车辆的转向意图的精准度，是需要考虑的问题。In recent years, autonomous driving technology and advanced driver assistance systems have entered a period of rapid development, and the driving safety of autonomous vehicles on the road has therefore become the focus of attention. The safe driving of autonomous vehicles on the road requires autonomous vehicles to accurately perceive, understand and predict the driving trajectories or steering intentions of various traffic participants in order to make safe and reasonable obstacle avoidance decisions. The accuracy of predicting the steering intention of the target vehicle is related to the driving safety of the autonomous vehicle on the road. How to improve the accuracy of predicting the steering intention of the target vehicle is a problem that needs to be considered.

发明内容Contents of the invention

本申请提供一种目标车辆转向意图的确定方法及装置，以期提高预测目标车辆转向意图的精准度，从而辅助自动驾驶，保证自动驾驶车辆在道路上的行驶安全性。The present application provides a method and device for determining the steering intention of a target vehicle, in order to improve the accuracy of predicting the steering intention of the target vehicle, thereby assisting automatic driving and ensuring the driving safety of the automatic driving vehicle on the road.

第一方面，提供一种目标车辆转向意图的确定方法，该方法包括以下步骤：获取目标车辆的视觉信息；根据视觉信息确定目标车辆的姿态信息，其中姿态信息为表示目标车辆的车身姿态的信息；根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图。In a first aspect, a method for determining a steering intention of a target vehicle is provided, the method comprising the following steps: acquiring visual information of the target vehicle; determining attitude information of the target vehicle according to the visual information, wherein the attitude information is information representing the body attitude of the target vehicle ; Determine the steering intention of the target vehicle according to the attitude information of the target vehicle.

根据目标车辆的姿态信息来预估目标车辆的转向意图，易于实现，车辆避免了机器学习方法对于训练数据和道路环境的要求，该方法适用范围更广。目标车辆的姿态信息能够更精准的表示目标车辆的转向意图，通过该方法所确定的目标车辆的转向意图能够更加精确，有助于提高预测目标车辆的转向意图的精确度，更好的辅助自动驾驶，保证自动驾驶车辆在道路上的行驶安全性。在一个可能的设计中，姿态信息包括目标车辆的侧倾角。目标车辆的侧倾角与目标车辆的横向加速度有直接联系，目标车辆的横向加速度能够表征目标车辆的转向意图，从而可知，根据目标车辆的侧倾角能够确定目标车辆的转向意图。通过目标车辆的侧倾角确定目标车辆的转向意图，提高预测精度，以更好的提高自动驾驶车辆的安全驾驶性能。Estimating the steering intention of the target vehicle based on the attitude information of the target vehicle is easy to implement, and the vehicle avoids the requirements of the machine learning method for training data and road environment, and the method is applicable to a wider range. The attitude information of the target vehicle can more accurately represent the steering intention of the target vehicle, and the steering intention of the target vehicle determined by this method can be more accurate, which helps to improve the accuracy of predicting the steering intention of the target vehicle, and better assists automatic Driving to ensure the safety of self-driving vehicles on the road. In one possible design, the attitude information includes the roll angle of the target vehicle. The roll angle of the target vehicle is directly related to the lateral acceleration of the target vehicle, and the lateral acceleration of the target vehicle can represent the steering intention of the target vehicle, so it can be known that the steering intention of the target vehicle can be determined according to the roll angle of the target vehicle. Determine the steering intention of the target vehicle through the roll angle of the target vehicle, improve the prediction accuracy, and better improve the safe driving performance of the autonomous vehicle.

在一个可能的设计中，根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图，通过以下方式实现：当所述侧倾角小于第一阈值时，确定所述目标车辆的转向意图为左转；或者，当所述侧倾角大于所述第一阈值小于第二阈值时，确定所述目标车辆的转向意图为直行；或者，当所述侧倾角大于所述第二阈值时，确定所述目标车辆的转向意图为右转；其中，所述第一阈值为负值，所述第二阈值为正值。可选的，第一阈值和第二阈值都为正值。In a possible design, according to the attitude information of the target vehicle, the steering intention of the target vehicle is determined in the following manner: when the roll angle is smaller than a first threshold, it is determined that the steering intention of the target vehicle is turn left; or, when the roll angle is greater than the first threshold and less than the second threshold, determine that the target vehicle’s steering intention is to go straight; or, when the roll angle is greater than the second threshold, determine that the The steering intention of the target vehicle is to turn right; wherein, the first threshold is a negative value, and the second threshold is a positive value. Optionally, both the first threshold and the second threshold are positive values.

在一个可能的设计中，根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图，包括：当侧倾角大于第一阈值小于与90°时，确定目标车辆的转向意图为左转；或者，当侧倾角大于90°小于第二阈值时，确定目标车辆的转向意图为直行；或者，当侧倾角大于第二阈值时，确定目标车辆的转向意图为右转。In a possible design, determining the steering intention of the target vehicle according to the attitude information of the target vehicle includes: determining that the steering intention of the target vehicle is a left turn when the roll angle is greater than a first threshold and less than 90°; Or, when the roll angle is greater than 90° and smaller than the second threshold, it is determined that the target vehicle's steering intention is to go straight; or, when the roll angle is greater than the second threshold, it is determined that the target vehicle's steering intention is to turn right.

在一个可能的设计中，根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图，包括：当侧倾角介于(90°-第一阈值)与90°之间值时，确定目标车辆的转向意图为左转。或者，当侧倾角大于(90°-第一阈值)小于(90°+第二阈值)时，确定目标车辆的转向意图为直行。或者，当侧倾角大于(90°+第二阈值)时，确定目标车辆的转向意图为右转。In a possible design, determining the steering intention of the target vehicle according to the attitude information of the target vehicle includes: when the roll angle is between (90°-the first threshold) and 90°, determining the target The steering intent of the vehicle is to turn left. Alternatively, when the roll angle is greater than (90°-the first threshold) and less than (90°+the second threshold), it is determined that the steering intention of the target vehicle is to go straight. Or, when the roll angle is greater than (90°+second threshold), it is determined that the steering intention of the target vehicle is a right turn.

在一个可能的设计中，所述目标车辆的侧倾角通过以下参数确定：所述目标车辆的空间法向量、以及地面法向量。In a possible design, the roll angle of the target vehicle is determined by the following parameters: a space normal vector of the target vehicle and a ground normal vector.

在一个可能的设计中，所述侧倾角符合以下公式：In one possible design, the roll angle conforms to the following formula:

其中，

为所述侧倾角，n _r为所述目标车辆的空间法向量，n _g为所述地面法向量。

in,

is the roll angle, n _r is the space normal vector of the target vehicle, and n _g is the ground normal vector.

在一个可能的设计中，所述目标车辆的姿态信息还包括转向角；所述目标车辆的车轮包括前车轮和后车轮；所述转向角为所述前车轮的空间法向量与所述后车轮的空间法向量之间的夹角。In a possible design, the attitude information of the target vehicle also includes a steering angle; the wheels of the target vehicle include front wheels and rear wheels; the steering angle is the space normal vector of the front wheel and the rear wheel The angle between the space normal vectors of .

在一个可能的设计中，所述转向角符合下述公式：In a possible design, the steering angle conforms to the following formula:

其中，δ为所述转向角，n _f为所述前车轮的空间法向量；n _r为所述后车轮的空间法向量。

Wherein, δ is the steering angle, n _f is the space normal vector of the front wheel; n _r is the space normal vector of the rear wheel.

在一个可能的设计中，根据所述视觉信息确定所述目标车辆的姿态信息，可以通过以下方式实现：确定所述目标车辆在所述视觉信息中对应的几何形状；根据所述几何形状确定所述目标车辆的姿态信息。In a possible design, determining the attitude information of the target vehicle according to the visual information may be achieved in the following manner: determining the corresponding geometric shape of the target vehicle in the visual information; determining the attitude information of the target vehicle according to the geometric shape Describe the attitude information of the target vehicle.

目标车辆在视觉信息中对应的几何形状可以是目标车辆的车轮在视觉信息中对应的第一几何形状，确定目标车辆的车轮在视觉信息中对应的第一几何形状。例如，目标车辆为直排轮车，直排轮车的车轮在视觉信息中对应的第一几何形状为椭圆。The geometric shape corresponding to the target vehicle in the visual information may be a first geometric shape corresponding to the wheels of the target vehicle in the visual information, and the first geometric shape corresponding to the wheels of the target vehicle in the visual information is determined. For example, the target vehicle is a straight-wheel vehicle, and the first geometric shape corresponding to the wheel of the straight-wheel vehicle in the visual information is an ellipse.

假设目标车辆的车轮在视觉信息中的几何形状为椭圆，将二维图像中的椭圆转换到三维相机坐标系中。该步骤可以通过以下方式实现：利用相机投影模型将椭圆转换为三维相机坐标系中的椭圆锥面。利用空间圆姿态测量方法从椭圆锥面中确定第一平面，第一平面的法向量即目标车辆的车轮在三维相机坐标系中所在平面的法向量。从二维图像即可获取目标车辆的三维运动姿态，减少了自动驾驶车辆对于激光雷达和毫米波雷达的依赖。Assuming that the geometric shape of the wheel of the target vehicle in the visual information is an ellipse, the ellipse in the 2D image is transformed into the 3D camera coordinate system. This step can be realized in the following way: using the camera projection model to transform the ellipse into an ellipse cone in the three-dimensional camera coordinate system. The first plane is determined from the ellipse cone by using the spatial circular attitude measurement method, and the normal vector of the first plane is the normal vector of the plane where the wheels of the target vehicle are located in the three-dimensional camera coordinate system. The three-dimensional motion posture of the target vehicle can be obtained from the two-dimensional image, which reduces the dependence of autonomous vehicles on lidar and millimeter-wave radar.

可选的，第一平面可以是目标车辆的后车轮所在的平面，也可以是目标车辆的车架所在的平面。目标车辆的后车轮与车架在同一个平面。假设垂直于地面的平面为第二平面，第一平面与第二平面的交线为第一平面与地面的交线，则第一平面与第二平面之间的夹角可以认为是侧倾角。Optionally, the first plane may be the plane where the rear wheels of the target vehicle are located, or the plane where the frame of the target vehicle is located. The rear wheels of the target vehicle are in the same plane as the frame. Assuming that the plane perpendicular to the ground is the second plane, and the intersection line between the first plane and the second plane is the intersection line between the first plane and the ground, the angle between the first plane and the second plane can be considered as the roll angle.

在一个可能的设计中，当目标车辆的车轮在视觉信息中的第一几何形状为椭圆时，还可以包括下述步骤：确定所述视觉信息中的多个椭圆；按照约束条件在所述多个椭圆中确定所述目标车辆的车轮在所述视觉信息中对应的第一椭圆，所述约定条件包括：所述椭圆在所述视觉信息中的尺寸占比满足设定的比例范围。也即是，第一椭圆为在所述视觉信息中的多个椭圆中、尺寸占比满足设定的比例范围的椭圆，所述第一椭圆为所述目标车辆的车轮在所述视觉信息中的椭圆。这样可以去除一些过大或过小的椭圆，去除非车轮对应的造成椭圆。该设定的比例范围例如可以是1/3、或2/5，或其他合理的比例。In a possible design, when the first geometric shape of the wheel of the target vehicle in the visual information is an ellipse, the following steps may also be included: determining a plurality of ellipses in the visual information; Determine the first ellipse corresponding to the wheel of the target vehicle in the visual information among the ellipses, and the agreed condition includes: the size ratio of the ellipse in the visual information satisfies a set ratio range. That is to say, the first ellipse is an ellipse whose size ratio satisfies a set ratio range among the multiple ellipses in the visual information, and the first ellipse is the wheel of the target vehicle in the visual information. ellipse. In this way, some ellipses that are too large or too small can be removed, and ellipses that are not corresponding to the wheel can be removed. The set ratio range may be, for example, 1/3, or 2/5, or other reasonable ratios.

在一个可能的设计中，所述约定条件还包括：所述椭圆的中心位置在所述视觉信息的中心线的下方，所述中心线将所述视觉信息分为上下两部分，可选的，所述中心线将所述视觉信息分为均匀的上下两部分。目标车辆的前车轮会在边界框的左下方，后车轮会在边界框的右下方，因此，将椭圆的中心位置约束在中心线的下方，能够去除一些由背景造成的椭圆噪声。In a possible design, the agreed condition further includes: the center position of the ellipse is below the center line of the visual information, and the center line divides the visual information into upper and lower parts. Optionally, The center line divides the visual information into even upper and lower parts. The front wheels of the target vehicle will be at the bottom left of the bounding box, and the rear wheels will be at the bottom right of the bounding box. Therefore, constraining the center position of the ellipse to be below the centerline can remove some ellipse noise caused by the background.

在一个可能的设计中，所述约定条件还包括：椭圆的边缘的颜色趋势符合：边缘外侧为深，边缘内侧为浅。也即是，所述目标车辆的车轮在所述视觉信息中对应的所述第一椭圆，该第一椭圆的边缘的颜色符合：边缘外侧的颜色深于边缘内侧的颜色。通过“外深内浅”的极性约束，能够选择轮胎与轮毂交界的边沿对应的椭圆。In a possible design, the agreed condition further includes: the color trend of the edge of the ellipse conforms to: the outer edge is dark, and the inner edge is light. That is, for the first ellipse corresponding to the wheel of the target vehicle in the visual information, the color of the edge of the first ellipse conforms to: the color outside the edge is darker than the color inside the edge. Through the polarity constraint of "deep on the outside and shallow on the inside", the ellipse corresponding to the border between the tire and the hub can be selected.

在一个可能的设计中，所述视觉信息为包括所述目标车辆的边界框。这样，通过可以检测边界框中目标车辆的姿态信息，可以排除除目标车辆之外的其余事物的干扰，更利于检测目标车辆的姿态信息。In a possible design, the visual information is a bounding box including the target vehicle. In this way, by being able to detect the attitude information of the target vehicle in the bounding box, the interference of other things except the target vehicle can be eliminated, which is more conducive to detecting the attitude information of the target vehicle.

在一个可能的设计中，所述目标车辆的姿态信息还包括车身长度，车身长度可以为目标车辆的轴距；例如，所述目标车辆的车轮包括前车轮和后车轮；所述轴距为所述前车轮和所述后车轮的空间距离。可以根据目标车辆的车身长度，确定目标车辆的转向意图。In a possible design, the attitude information of the target vehicle also includes the body length, which can be the wheelbase of the target vehicle; for example, the wheels of the target vehicle include front wheels and rear wheels; the wheelbase is the The space distance between the front wheel and the rear wheel. The steering intention of the target vehicle can be determined according to the body length of the target vehicle.

转向角、轴距或两者的结合，可以用于确定目标车辆的转弯半径，从而根据目标车辆的转弯半径，确定安全驾驶策略。Steering angle, wheelbase, or a combination of both can be used to determine the turning radius of the target vehicle, so that a safe driving strategy can be determined based on the turning radius of the target vehicle.

可选的，上述目标车辆可以为直排轮车，例如，自行车、电动车或燃油车。其中，电动车为直排轮的电瓶车，燃油车例如可以是摩托车。Optionally, the above-mentioned target vehicle may be a straight-wheel vehicle, for example, a bicycle, an electric vehicle or a fuel vehicle. Wherein, the electric vehicle is a storage battery vehicle with inline wheels, and the fuel vehicle may be a motorcycle, for example.

在一个可能的设计中，根据所述目标车辆的转向意图，确定安全驾驶策略，即确定安全合理的自动驾驶策略，能够提高自动驾驶车辆在路上行驶的安全性。示例地，执行该转向意图确定方法的终端，可以相应的采取避开目标车辆的策略，如控制该终端停止避开目标车辆，或者控制该终端转向避开目标车辆等。其中，本申请的终端除了包括车辆以外，还可以包括用户设备、接入终端、用户单元、用户站、移动站、移动台、远方站、远程终端、移动设备、用户终端、终端设备、无线通信设备、用户代理或用户装置。终端设备还可以是蜂窝电话、无绳电话、SIP电话、WLL站、PDA、具有无线通信功能的手持设备、计算设备或连接到无线调制解调器的其它处理设备、车载设备、可穿戴设备、智能家居设备，智能机器人等，本申请实施例对此并不限定。In a possible design, according to the steering intention of the target vehicle, a safe driving strategy is determined, that is, a safe and reasonable automatic driving strategy is determined, which can improve the safety of the automatic driving vehicle on the road. For example, the terminal executing the method for determining the turning intention may adopt a corresponding strategy to avoid the target vehicle, such as controlling the terminal to stop avoiding the target vehicle, or controlling the terminal to turn to avoid the target vehicle. Wherein, besides the vehicle, the terminal in the present application may also include user equipment, access terminal, subscriber unit, subscriber station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal equipment, wireless communication device, user agent, or user device. The terminal device can also be a cellular phone, a cordless phone, a SIP phone, a WLL station, a PDA, a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a wearable device, a smart home device, Intelligent robots and the like are not limited in this embodiment of the present application.

本申请实施例中，获取目标车辆的视觉信息；检测视觉信息中目标车辆的姿态信息，根据目标车辆的姿态信息，确定目标车辆的转向意图。还可以扩展该方案为：获取驾驶目标车辆的人的视觉信息；检测视觉信息中所述人的肢体的姿态信息，根据所述人的肢体的姿态信息，确定目标车辆的转向意图。In the embodiment of the present application, the visual information of the target vehicle is obtained; the posture information of the target vehicle in the visual information is detected, and the steering intention of the target vehicle is determined according to the posture information of the target vehicle. The solution can also be extended to: acquire the visual information of the person driving the target vehicle; detect the posture information of the human limbs in the visual information, and determine the steering intention of the target vehicle according to the posture information of the human limbs.

第二方面，本申请提供一种车辆转向意图的确定装置，该装置包括处理模块和输入输出模块。输入输出模块可用于获取目标车辆的视觉信息；处理模块，用于根据所述视觉信息确定目标车辆的姿态信息，其中姿态信息为表示目标车辆的车身姿态的信息；以及用于根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图。In a second aspect, the present application provides a device for determining a steering intention of a vehicle, and the device includes a processing module and an input-output module. The input-output module can be used to obtain the visual information of the target vehicle; the processing module is used to determine the attitude information of the target vehicle according to the visual information, wherein the attitude information is information representing the body attitude of the target vehicle; The attitude information of the target vehicle is determined to determine the steering intention of the target vehicle.

在一个可能的设计中，姿态信息包括目标车辆的侧倾角。In one possible design, the attitude information includes the roll angle of the target vehicle.

在一个可能的设计中，在根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图时，处理模块用于：当所述侧倾角小于第一阈值时，确定所述目标车辆的转向意图为左转；或者，当所述侧倾角大于所述第一阈值小于第二阈值时，确定所述目标车辆的转向意图为直行；或者，当所述侧倾角大于所述第二阈值时，确定所述目标车辆的转向意图为右转；其中，所述第一阈值为负值，所述第二阈值为正值。可选的，第一阈值和第二阈值都为正值。In a possible design, when determining the steering intention of the target vehicle according to the attitude information of the target vehicle, the processing module is configured to: determine the steering of the target vehicle when the roll angle is smaller than a first threshold The intention is to turn left; or, when the roll angle is greater than the first threshold and less than a second threshold, it is determined that the steering intention of the target vehicle is to go straight; or, when the roll angle is greater than the second threshold, It is determined that the steering intention of the target vehicle is a right turn; wherein, the first threshold is a negative value, and the second threshold is a positive value. Optionally, both the first threshold and the second threshold are positive values.

其中，

in,

在一个可能的设计中，在根据所述视觉信息确定所述目标车辆的姿态信息时，处理模块具体用于：确定所述目标车辆在所述视觉信息中对应的几何形状；根据所述几何形状确定所述目标车辆的姿态信息。In a possible design, when determining the attitude information of the target vehicle according to the visual information, the processing module is specifically configured to: determine the corresponding geometric shape of the target vehicle in the visual information; Determine the attitude information of the target vehicle.

假设目标车辆的车轮在视觉信息中的几何形状为椭圆，将二维图像中的椭圆转换到三维相机坐标系中第一平面或第一平面的法向量。处理模块具体执行以下操作：利用相机投影模型将椭圆转换为三维相机坐标系中的椭圆锥面。利用空间圆姿态测量方法从椭圆锥面中确定第一平面，第一平面的法向量即目标车辆的车轮在三维相机坐标系中所在平面的法向量。从二维图像即可获取目标车辆的三维运动姿态，减少了自动驾驶车辆对于激光雷达和毫米波雷达的依赖。Assuming that the geometric shape of the wheel of the target vehicle in the visual information is an ellipse, the ellipse in the 2D image is converted to the first plane or the normal vector of the first plane in the 3D camera coordinate system. The processing module specifically performs the following operations: transform the ellipse into an ellipse cone in the three-dimensional camera coordinate system by using the camera projection model. The first plane is determined from the ellipse cone by using the spatial circular attitude measurement method, and the normal vector of the first plane is the normal vector of the plane where the wheels of the target vehicle are located in the three-dimensional camera coordinate system. The three-dimensional motion posture of the target vehicle can be obtained from the two-dimensional image, which reduces the dependence of autonomous vehicles on lidar and millimeter-wave radar.

在一个可能的设计中，当目标车辆的车轮在视觉信息中的第一几何形状为椭圆时，处理模块还用于：确定所述视觉信息中的多个椭圆；按照约束条件在所述多个椭圆中确定所述目标车辆的车轮在所述视觉信息中对应的第一椭圆，所述约定条件包括：所述椭圆在所述视觉信息中的尺寸占比满足设定的比例范围。也即是，第一椭圆为在所述视觉信息中的多个椭圆中、尺寸占比满足设定的比例范围的椭圆，所述第一椭圆为所述目标车辆的车轮在所述视觉信息中的椭圆。这样可以去除一些过大或过小的椭圆，去除非车轮对应的造成椭圆。该设定的比例范围例如可以是1/3、或2/5，或其他合理的比例。In a possible design, when the first geometric shape of the wheel of the target vehicle in the visual information is an ellipse, the processing module is further configured to: determine a plurality of ellipses in the visual information; The first ellipse corresponding to the wheel of the target vehicle in the visual information is determined among the ellipses, and the agreed condition includes: the proportion of the size of the ellipse in the visual information satisfies a set ratio range. That is to say, the first ellipse is an ellipse whose size ratio satisfies a set ratio range among the multiple ellipses in the visual information, and the first ellipse is the wheel of the target vehicle in the visual information. ellipse. In this way, some ellipses that are too large or too small can be removed, and ellipses that are not corresponding to the wheels can be removed. The set ratio range may be, for example, 1/3, or 2/5, or other reasonable ratios.

在一个可能的设计中，所述目标车辆的姿态信息还包括车身长度，车身长度可以为目标车辆的轴距；例如，所述目标车辆的车轮包括前车轮和后车轮；所述轴距为所述前车轮和所述后车轮的空间距离。处理模块还用于可以根据目标车辆的车身长度，确定目标车辆的转向意图。In a possible design, the attitude information of the target vehicle also includes the body length, which can be the wheelbase of the target vehicle; for example, the wheels of the target vehicle include front wheels and rear wheels; the wheelbase is the The space distance between the front wheel and the rear wheel. The processing module is also used to determine the steering intention of the target vehicle according to the body length of the target vehicle.

转向角、轴距、或两者的结合，可以用于确定目标车辆的转弯半径，从而根据目标车辆的转弯半径，确定安全驾驶策略。Steering angle, wheelbase, or the combination of both can be used to determine the turning radius of the target vehicle, so as to determine a safe driving strategy according to the turning radius of the target vehicle.

在一个可能的设计中，处理模块具体用于根据所述目标车辆的转向意图，确定安全驾驶策略，即确定安全合理的自动驾驶策略，能够提高自动驾驶车辆在路上行驶的安全性。示例地，执行该转向意图确定方法的终端，可以相应的采取避开目标车辆的策略，如控制该终端停止避开目标车辆，或者控制该终端转向避开目标车辆等。其中，本申请的终端除了包括车辆以外，还可以包括用户设备、接入终端、用户单元、用户站、移动站、移动台、远方站、远程终端、移动设备、用户终端、终端设备、无线通信设备、用户代理或用户装置。终端设备还可以是蜂窝电话、无绳电话、会话启动协议(session initiation protocol，SIP)电话、无线本地环路(wireless local loop，WLL)站、个人数字助理(personal digital assistant，PDA)、具有无线通信功能的手持设备、计算设备或连接到无线调制解调器的其它处理设备、车载设备、可穿戴设备、智能家居设备，智能机器人等，本申请实施例对此并不限定。In a possible design, the processing module is specifically configured to determine a safe driving strategy according to the steering intention of the target vehicle, that is, to determine a safe and reasonable automatic driving strategy, which can improve the safety of the automatic driving vehicle on the road. For example, the terminal executing the method for determining the turning intention may adopt a corresponding strategy to avoid the target vehicle, such as controlling the terminal to stop avoiding the target vehicle, or controlling the terminal to turn to avoid the target vehicle. Wherein, besides the vehicle, the terminal in the present application may also include user equipment, access terminal, subscriber unit, subscriber station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal equipment, wireless communication device, user agent, or user device. The terminal device can also be a cellular phone, a cordless phone, a session initiation protocol (session initiation protocol, SIP) phone, a wireless local loop (wireless local loop, WLL) station, a personal digital assistant (personal digital assistant, PDA), a Functional handheld devices, computing devices or other processing devices connected to wireless modems, vehicle-mounted devices, wearable devices, smart home devices, smart robots, etc., are not limited in this embodiment of the present application.

第二方面的有益效果可以参考第一方面对应的描述，在此不再赘述。For the beneficial effects of the second aspect, reference may be made to the corresponding description of the first aspect, and details are not repeated here.

第三方面，本申请提供一种计算设备，包括处理器，处理器与存储器相连，存储器存储计算机程序或指令，处理器用于执行存储器中存储的计算机程序或指令，以使得计算设备执行上述第一方面或第一方面的任一种可能的实现方式中的方法。In a third aspect, the present application provides a computing device, including a processor, the processor is connected to a memory, the memory stores computer programs or instructions, and the processor is used to execute the computer programs or instructions stored in the memory, so that the computing device performs the above-mentioned first Aspect or a method in any possible implementation of the first aspect.

第四方面，本申请提供一种计算机可读存储介质，其上存储有计算机程序或指令，当该计算机程序或指令被执行时，使得计算机执行上述第一方面或第一方面的任一种可能的实现方式中的方法。In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program or instruction is stored, and when the computer program or instruction is executed, the computer executes any one of the above-mentioned first aspect or the first aspect. method in the implementation of .

第五方面，本申请提供一种计算机程序产品，当计算机执行计算机程序产品时，使得计算机执行上述第一方面或第一方面的任一种可能的实现方式中的方法。In a fifth aspect, the present application provides a computer program product. When a computer executes the computer program product, the computer executes the method in the above-mentioned first aspect or any possible implementation manner of the first aspect.

第六方面，本申请提供一种芯片，芯片与存储器相连，用于读取并执行存储器中存储的计算机程序或指令，以实现上述第一方面或第一方面的任一种可能的实现方式中的方法。In a sixth aspect, the present application provides a chip connected to a memory for reading and executing computer programs or instructions stored in the memory, so as to realize the above-mentioned first aspect or any possible implementation of the first aspect Methods.

第七方面，本申请提供一种自动驾驶车辆，该自动驾驶车辆包括上述第二方面或第二方面的任一种可能的实现方式中的目标车辆转向意图的确定装置和执行装置，以实现上述第一方面或第一方面的任一种可能的实现方式中的方法。In a seventh aspect, the present application provides an automatic driving vehicle, which includes the device for determining and executing the steering intention of the target vehicle in the second aspect or any possible implementation of the second aspect, so as to realize the above-mentioned The first aspect or the method in any possible implementation of the first aspect.

第八方面，本申请提供一种自动驾驶车辆，该自动驾驶车辆包括上述第六方面中的芯片和执行装置，以实现上述第一方面或第一方面的任一种可能的实现方式中的方法。In an eighth aspect, the present application provides an autonomous vehicle, which includes the chip and execution device in the sixth aspect above, so as to implement the method in the first aspect or any possible implementation of the first aspect .

第九方面，本申请还提供一种安全驾驶方法，该方法可以由自动驾驶车辆执行，包括以下步骤，根据如上述第一方面或第一方面任何一种可能的设计所述的目标车辆转向意图的确定方法，确定安全驾驶策略。In the ninth aspect, the present application also provides a safe driving method, which can be executed by an autonomous vehicle, including the following steps, according to the steering intention of the target vehicle as described in the first aspect or any possible design of the first aspect The determination method to determine the safe driving strategy.

附图说明Description of drawings

图1为本申请实施例中自动驾驶车辆的结构示意图；FIG. 1 is a schematic structural diagram of a self-driving vehicle in an embodiment of the present application;

图2a为本申请实施例中直排轮车的示意图之一；Fig. 2 a is one of the schematic diagrams of the inline-wheeled vehicle in the embodiment of the present application;

图2b为本申请实施例中直排轮车的示意图之二；Fig. 2b is the second schematic diagram of the straight-wheeled vehicle in the embodiment of the present application;

图2c为本申请实施例中直排轮车的示意图之二；Fig. 2c is the second schematic diagram of the straight-wheeled vehicle in the embodiment of the present application;

图3为本申请实施例中应用场景示意图；FIG. 3 is a schematic diagram of an application scenario in an embodiment of the present application;

图4为本申请实施例中目标车辆转向意图的确定方法的流程示意图；FIG. 4 is a schematic flowchart of a method for determining a target vehicle's steering intention in an embodiment of the present application;

图5为本申请实施例中边界框示意图；FIG. 5 is a schematic diagram of a bounding box in an embodiment of the present application;

图6为本申请实施例中第一平面的求解示意图；Fig. 6 is a schematic diagram of solving the first plane in the embodiment of the present application;

图7为本申请实施例中椭圆去噪声的示意图；FIG. 7 is a schematic diagram of ellipse denoising in the embodiment of the present application;

图8为本申请实施例中姿态信息的示意图；FIG. 8 is a schematic diagram of posture information in an embodiment of the present application;

图9a为本申请实施例中第一平面、与第二平面之间的关系示意图之一；Fig. 9a is one of the schematic diagrams of the relationship between the first plane and the second plane in the embodiment of the present application;

图9b为本申请实施例中第一平面、与第二平面之间的关系示意图之二；Fig. 9b is the second schematic diagram of the relationship between the first plane and the second plane in the embodiment of the present application;

图10为本申请实施例中目标车辆的车身姿态的视觉信息示意图；FIG. 10 is a schematic diagram of the visual information of the body posture of the target vehicle in the embodiment of the present application;

图11为本申请实施例中自动驾驶车辆的模块示意图；Fig. 11 is a schematic diagram of modules of the self-driving vehicle in the embodiment of the present application;

图12为本申请实施例中目标车辆转向意图的确定装置结构示意图之一；FIG. 12 is one of the structural schematic diagrams of the device for determining the steering intention of the target vehicle in the embodiment of the present application;

图13为本申请实施例中目标车辆转向意图的确定装置结构示意图之二。FIG. 13 is the second structural schematic diagram of the device for determining the steering intention of the target vehicle in the embodiment of the present application.

具体实施方式Detailed ways

本申请提供一种目标车辆转向意图的确定方法及装置，以期提高预测目标车辆转向意图的精准度，从而辅助自动驾驶，保证自动驾驶车辆在道路上的行驶安全性。其中，方法和装置是基于同一技术构思的，由于方法及装置解决问题的原理相似，因此装置与方法的实施可以相互参见，重复之处不再赘述。The present application provides a method and device for determining the steering intention of a target vehicle, in order to improve the accuracy of predicting the steering intention of the target vehicle, thereby assisting automatic driving and ensuring the driving safety of the automatic driving vehicle on the road. Among them, the method and the device are based on the same technical conception. Since the principle of solving the problem of the method and the device is similar, the implementation of the device and the method can be referred to each other, and the repetition will not be repeated.

下面将结合附图，对本申请实施例进行详细描述。Embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

本申请也可适用于高级驾驶辅助***(advanced driving assistant system，ADAS)，机器人、无人机、网联车、安防监控等领域。ADAS例如可以是自动驾驶。本申请可适用于自动驾驶车辆或集成了ADAS的车辆，例如可以是具有人机交互(human machine interaction，HMI)功能的自动驾驶车辆，可以是针对自动驾驶车辆的驾驶状态通过自动驾驶算法进行计算和判断功能的自动驾驶车辆，也可以是对自动驾驶车辆进行运动控制功能的自动驾驶车辆。This application can also be applied to the fields of advanced driving assistant system (advanced driving assistant system, ADAS), robot, unmanned aerial vehicle, networked vehicle, security monitoring, etc. ADAS can be, for example, autonomous driving. This application can be applied to self-driving vehicles or vehicles integrated with ADAS, for example, it can be a self-driving vehicle with human machine interaction (HMI) function, and it can calculate the driving state of the self-driving vehicle through an automatic driving algorithm It can also be an automatic driving vehicle that performs a motion control function on the automatic driving vehicle.

可选的，自动驾驶车辆可以包含至少一个自动驾驶***，以支持自动驾驶车辆进行自动驾驶。Optionally, the self-driving vehicle may include at least one automatic driving system to support the automatic driving of the self-driving vehicle.

应理解，根据实际使用的需要，也可将自动驾驶车辆替换为火车、飞行器、机器人、慢速运输车或移动平台等其他载具或交通工具。本申请对此不做限定。It should be understood that, according to the needs of actual use, the self-driving vehicle can also be replaced by other vehicles or vehicles such as trains, aircraft, robots, slow transport vehicles or mobile platforms. This application does not limit this.

如图1所示，示出了本申请实施例中自动驾驶车辆的一种功能框图。在一个实施例中，自动驾驶车辆100可以配置为完全或部分地自动驾驶模式。例如，自动驾驶车辆100可以在处于自动驾驶模式中同时控制自身，并且可以通过人为操作来确定自动驾驶车辆及其周边环境的当前状态，确定周边环境中的至少一个其他自动驾驶车辆的可能行为，并确定该其他自动驾驶车辆执行可能行为的可能性相对应的置信水平，并基于所确定的信息来控制自动驾驶车辆100。在自动驾驶车辆100处于自动驾驶模式中时，可以将自动驾驶车辆100置为在没有和人交互的情况下操作。As shown in FIG. 1 , it shows a functional block diagram of the self-driving vehicle in the embodiment of the present application. In one embodiment, autonomous vehicle 100 may be configured in a fully or partially autonomous driving mode. For example, the self-driving vehicle 100 can control itself while being in the self-driving mode, and can determine the current state of the self-driving vehicle and its surrounding environment through human operations, determine the possible behavior of at least one other self-driving vehicle in the surrounding environment, A confidence level corresponding to the likelihood of the other autonomous vehicle performing the possible action is determined, and the autonomous vehicle 100 is controlled based on the determined information. While the self-driving vehicle 100 is in the self-driving mode, the self-driving vehicle 100 may be set to operate without human interaction.

如图1所示，耦合到自动驾驶车辆100或包括在自动驾驶车辆100中的组件可以包括推进***110、传感器***120、控制***130、***设备140、电源150、计算机***160以及用户接口170。自动驾驶车辆100的组件可以被配置为以与彼此互连和/或与耦合到各***的其它组件互连的方式工作。例如，电源150可以向自动驾驶车辆100的所有组件提供电力。计算机***160可以被配置为从推进***110、传感器***120、控制***130和***设备140接收数据并对它们进行控制。计算机***160还可以被配置为在用户接口170上生成图像的显示并从用户接口170接收输入。As shown in FIG. 1 , components coupled to or included in autonomous vehicle 100 may include propulsion system 110 , sensor system 120 , control system 130 , peripherals 140 , power supply 150 , computer system 160 , and user interface 170 . Components of autonomous vehicle 100 may be configured to work interconnected with each other and/or with other components coupled to various systems. For example, power supply 150 may provide power to all components of autonomous vehicle 100 . Computer system 160 may be configured to receive data from and control propulsion system 110 , sensor system 120 , control system 130 , and peripherals 140 . Computer system 160 may also be configured to generate a display of images on user interface 170 and to receive input from user interface 170 .

需要说明的是，在其它示例中，自动驾驶车辆100可以包括更多、更少或不同的***，并且每个***可以包括更多、更少或不同的组件。此外，示出的***和组件可以按任意种的方式进行组合或划分，本申请对此不做具体限定。It should be noted that, in other examples, the autonomous vehicle 100 may include more, fewer or different systems, and each system may include more, fewer or different components. In addition, the illustrated systems and components may be combined or divided in any manner, which is not specifically limited in the present application.

传感器***120可以包括用于感测自动驾驶车辆100周围环境的若干个传感器。如图1所示，传感器***120的传感器包括全球定位***(Global PositioningSystem，GPS)126、惯性测量单元(Inertial Measurement Unit，IMU)125、激光雷达传感器、相机传感器123、毫米波雷达传感器以及用于修改传感器的位置和/或朝向的制动器121。毫米波雷达传感器可利用无线电信号来感测自动驾驶车辆100的周边环境内的目标。在一些实施例中，除了感测目标以外，毫米波雷达122还可用于感测目标的速度和/或前进方向。激光雷达124可利用激光来感测自动驾驶车辆100所位于的环境中的目标。在一些实施例中，激光雷达124 可包括一个或多个激光源、激光扫描器以及一个或多个检测器，以及其他***组件。相机传感器123可用于捕捉自动驾驶车辆100的周边环境的多个图像。相机传感器123可以是静态相机或视频相机。The sensor system 120 may include several sensors for sensing the environment around the autonomous vehicle 100 . As shown in FIG. 1 , the sensors of the sensor system 120 include a global positioning system (Global Positioning System, GPS) 126, an inertial measurement unit (Inertial Measurement Unit, IMU) 125, a lidar sensor, a camera sensor 123, a millimeter wave radar sensor, and a Actuator 121 that modifies the position and/or orientation of the sensor. The millimeter wave radar sensor may utilize radio signals to sense objects within the surrounding environment of the autonomous vehicle 100 . In some embodiments, millimeter wave radar 122 may be used to sense the speed and/or heading of a target in addition to sensing the target. Lidar 124 may utilize laser light to sense objects in the environment in which autonomous vehicle 100 is located. In some embodiments, lidar 124 may include one or more laser sources, a laser scanner, and one or more detectors, among other system components. Camera sensor 123 may be used to capture multiple images of the surrounding environment of autonomous vehicle 100 . The camera sensor 123 may be a still camera or a video camera.

控制***130为控制自动驾驶车辆100及其组件的操作。控制***130可包括各种元件，其中包括转向单元136、油门135、制动单元134、传感器融合算法133、计算机视觉***132、路线控制***131以及障碍物避免***137。转向***136可操作来调整自动驾驶车辆100的前进方向。例如在一个实施例中可以为方向盘***。油门135用于控制引擎114的操作速度并进而控制自动驾驶车辆100的速度。控制***130可以额外地或可替换地包括除了图1所示出的组件以外的其他组件。本申请对此不做具体限定。The control system 130 controls the operation of the autonomous vehicle 100 and its components. Control system 130 may include various elements including steering unit 136 , accelerator 135 , braking unit 134 , sensor fusion algorithm 133 , computer vision system 132 , route control system 131 , and obstacle avoidance system 137 . Steering system 136 is operable to adjust the heading of autonomous vehicle 100 . For example in one embodiment it could be a steering wheel system. The throttle 135 is used to control the operating speed of the engine 114 and thus the speed of the autonomous vehicle 100 . Control system 130 may additionally or alternatively include other components than those shown in FIG. 1 . This application does not specifically limit it.

计算机视觉***132可以操作来处理和分析由相机传感器123捕捉的图像以便识别自动驾驶车辆100周边环境中的目标和/或特征。所述目标和/或特征可包括交通信号、道路边界和障碍物。计算机视觉***132可使用目标识别算法、运动中恢复结构(structure from motion，SFM)算法、视频跟踪和其他计算机视觉技术。在一些实施例中，计算机视觉***132可以用于为环境绘制地图、跟踪目标、估计目标的速度等等。路线控制***134用于确定自动驾驶车辆100的行驶路线。在一些实施例中，路线控制***142可结合来自传感器***120、GPS 126和一个或多个预定地图的数据以为自动驾驶车辆100确定行驶路线。障碍物避免***137用于识别、评估和避免或者以其他方式越过自动驾驶车辆100的环境中的潜在障碍物。当然，在一个实例中，控制***130可以增加或替换地包括除了所示出和描述的那些以外的组件。或者也可以减少一部分上述示出的组件。Computer vision system 132 may operate to process and analyze images captured by camera sensor 123 in order to identify objects and/or features in the environment surrounding autonomous vehicle 100 . The objects and/or features may include traffic signals, road boundaries and obstacles. The computer vision system 132 may use object recognition algorithms, structure from motion (SFM) algorithms, video tracking, and other computer vision techniques. In some embodiments, computer vision system 132 may be used to map the environment, track objects, estimate the speed of objects, and the like. The route control system 134 is used to determine the driving route of the autonomous vehicle 100 . In some embodiments, route control system 142 may combine data from sensor system 120, GPS 126, and one or more predetermined maps to determine a travel route for autonomous vehicle 100. Obstacle avoidance system 137 is used to identify, evaluate, and avoid or otherwise overcome potential obstacles in the environment of autonomous vehicle 100 . Of course, in one example, control system 130 may additionally or alternatively include components other than those shown and described. Alternatively, some of the components shown above may be reduced.

***设备140可以被配置为允许自动驾驶车辆100与外部传感器、其它自动驾驶车辆和/或用户交互。为此，***设备140可以包括例如无线通信***144、触摸屏143、麦克风142和/或扬声器141。***设备140可以额外地或可替换地包括除了图1所示出的组件以外的其他组件。本申请对此不做具体限定。Peripherals 140 may be configured to allow autonomous vehicle 100 to interact with external sensors, other autonomous vehicles, and/or a user. To this end, peripherals 140 may include, for example, a wireless communication system 144 , a touch screen 143 , a microphone 142 and/or a speaker 141 . Peripheral device 140 may additionally or alternatively include other components than those shown in FIG. 1 . This application does not specifically limit it.

电源150可以被配置为向自动驾驶车辆100的一些或全部组件提供电力。自动驾驶车辆100的组件可以被配置为以与在其各自的***内部和/或外部的其它组件互连的方式工作。为此，自动驾驶车辆100的组件和***可以通过***总线、网络和/或其它连接机制通信地链接在一起。Power source 150 may be configured to provide power to some or all components of autonomous vehicle 100 . Components of autonomous vehicle 100 may be configured to work in an interconnected manner with other components within and/or external to their respective systems. To this end, the components and systems of autonomous vehicle 100 may be communicatively linked together via a system bus, network, and/or other connection mechanisms.

自动驾驶车辆100的部分或所有功能受计算机***160控制。计算机***160可包括至少一个处理器161，处理器161执行存储在例如存储器163这样的非暂态计算机可读介质中的指令1631。计算机***160还可以是采用分布式方式控制自动驾驶车辆100的个体组件或子***的多个计算设备。Some or all functions of autonomous vehicle 100 are controlled by computer system 160 . Computer system 160 may include at least one processor 161 executing instructions 1631 stored in a non-transitory computer-readable medium such as memory 163 . Computer system 160 may also be a plurality of computing devices that control individual components or subsystems of autonomous vehicle 100 in a distributed manner.

处理器161可以是任何常规的处理器，诸如商业可获得的中央处理器(central processing unit，CPU)。替选地，该处理器可以是诸如专用集成电路(application specific integrated circuits,ASIC)或其它基于硬件的处理器的专用设备。尽管图1功能性地图示了处理器、存储器、和在相同块中的计算机***160的其它元件，但是本领域的普通技术人员应该理解该处理器、计算机、或存储器实际上可以包括可以或者可以不存储在相同的物理外壳内的多个处理器、计算机、或存储器。例如，存储器可以是硬盘驱动器或位于不同于计算机***160的外壳内的其它存储介质。因此，对处理器或计算机的引用将被理解为包括对可以或者可以不并行操作的处理器或计算机或存储器的集合的引用。不同于使用单一的处理器来执行此处所描述的步骤，诸如转向组件和减速组件的一些组件每个都可以具有其自己的处理器，所述处理器只执行与特定于组件的功能相关的计算。Processor 161 may be any conventional processor, such as a commercially available central processing unit (CPU). Alternatively, the processor may be a dedicated device such as an application specific integrated circuit (ASIC) or other hardware-based processor. Although FIG. 1 functionally illustrates the processor, memory, and other elements of computer system 160 in the same block, those of ordinary skill in the art will appreciate that the processor, computer, or memory may actually include or may include Multiple processors, computers, or memory that are not stored in the same physical enclosure. For example, memory may be a hard drive or other storage medium located in a different housing than computer system 160 . Accordingly, references to a processor or computer are to be understood to include references to collections of processors or computers or memories that may or may not operate in parallel. Instead of using a single processor to perform the steps described herein, some components, such as the steering and deceleration components, may each have their own processor that only performs calculations related to component-specific functions .

在此处所描述的各个方面中，处理器可以位于远离该自动驾驶车辆并且与该自动驾驶车辆进行无线通信。在其它方面中，此处所描述的过程中的一些在布置于自动驾驶车辆内的处理器上执行而其它则由远程处理器执行，包括采取执行单一操纵的必要步骤。In various aspects described herein, the processor can be located remotely from the autonomous vehicle and be in wireless communication with the autonomous vehicle. In other aspects, some of the processes described herein are executed on a processor disposed within the autonomous vehicle while others are executed by a remote processor, including taking the necessary steps to perform a single maneuver.

在一些实施例中，存储器163可包含指令1631(例如，程序逻辑)，指令1631可被处理器161执行来执行自动驾驶车辆100的各种功能，包括以上描述的那些功能。存储器163也可包含额外的指令，包括向推进***110、传感器***120、控制***130和***设备140中的一个或多个发送数据、从其接收数据、与其交互，和/或对其进行控制的指令。In some embodiments, memory 163 may contain instructions 1631 (eg, program logic) executable by processor 161 to perform various functions of autonomous vehicle 100 , including those described above. Memory 163 may also contain additional instructions, including sending data to, receiving data from, interacting with, and/or controlling one or more of propulsion system 110, sensor system 120, control system 130, and peripherals 140 instructions.

除了指令1631以外，存储器163还可存储数据，例如道路地图、路线信息，自动驾驶车辆的位置、方向、速度以及其它这样的自动驾驶车辆数据，以及其他信息。这种信息可在自动驾驶车辆100在自主、半自主和/或手动模式中操作期间被自动驾驶车辆100和计算机***160使用。In addition to instructions 1631, memory 163 may also store data such as road maps, route information, the location, direction, speed, and other such autonomous vehicle data of the autonomous vehicle, among other information. Such information may be used by autonomous vehicle 100 and computer system 160 during operation of autonomous vehicle 100 in autonomous, semi-autonomous, and/or manual modes.

用户接口170，用于向自动驾驶车辆100的用户提供信息或从其接收信息。可选地，用户接口170可包括在***设备140的集合内的一个或多个输入/输出设备，例如无线通信***144、触摸屏143、麦克风142和扬声器141。User interface 170 for providing information to or receiving information from a user of autonomous vehicle 100 . Optionally, user interface 170 may include one or more input/output devices within set of peripheral devices 140 , such as wireless communication system 144 , touch screen 143 , microphone 142 and speaker 141 .

计算机***160可基于从各种子***(例如，推进***110、传感器***120和控制***130)以及从用户接口170接收的输入来控制自动驾驶车辆100的功能。例如，计算机***160可利用来自控制***130的输入以便控制转向单元136来避免由传感器***120和障碍物避免***137检测到的障碍物。在一些实施例中，计算机***160可操作来对自动驾驶车辆100及其子***的许多方面提供控制。Computer system 160 may control functions of autonomous vehicle 100 based on input received from various subsystems (eg, propulsion system 110 , sensor system 120 , and control system 130 ) and from user interface 170 . For example, computer system 160 may utilize input from control system 130 in order to control steering unit 136 to avoid obstacles detected by sensor system 120 and obstacle avoidance system 137 . In some embodiments, computer system 160 is operable to provide control over many aspects of autonomous vehicle 100 and its subsystems.

可选地，上述这些组件中的一个或多个可与自动驾驶车辆100分开安装或关联。例如，存储器163可以部分或完全地与自动驾驶车辆100分开存在。上述组件可以按有线和/或无线方式来通信地耦合在一起。Optionally, one or more of these above-mentioned components may be separately installed or associated with the self-driving vehicle 100 . For example, the memory 163 may exist partially or completely separately from the autonomous vehicle 100 . The components described above may be communicatively coupled together in a wired and/or wireless manner.

可选地，上述组件只是一个示例，实际应用中，上述各个模块中的组件有可能根据实际需要增添或者删除，图1不应理解为对本申请实施例的限制。Optionally, the above-mentioned components are just an example. In practical applications, components in the above-mentioned modules may be added or deleted according to actual needs. FIG. 1 should not be construed as limiting the embodiment of the present application.

在道路行进自动驾驶的车辆，如上面的自动驾驶车辆100，可以识别其周围环境内的目标以确定对当前速度的调整。本申请实施例中，所述目标可以是直排轮车。在一些示例中，可以根据确定直排轮车的转向意图，用来确定自动驾驶车辆所要调整的速度。An autonomous vehicle traveling on a road, such as the autonomous vehicle 100 above, can identify objects within its surroundings to determine adjustments to the current speed. In this embodiment of the present application, the target may be a straight-wheeled vehicle. In some examples, the determination of the steering intent of the inline-wheel vehicle can be used to determine the speed at which the autonomous vehicle is to be adjusted.

可选地，自动驾驶车辆100或者与自动驾驶车辆100相关联的计算设备(如图1的计算机***160、计算机视觉***132、存储器163)可以基于所识别的目标的特性和周围环境的状态(例如，交通、雨、道路上的冰等等)来预测所述识别的目标的行为。可选地，每一个所识别的目标都依赖于彼此的行为，因此还可以将所识别的所有目标全部一起考虑来预测单个识别的目标的行为。自动驾驶车辆100能够基于预测的所述识别的目标的行为来调整它的速度。换句话说，自动驾驶汽车能够基于所预测的目标的行为来确定自动驾驶车辆将需要调整到(例如，加速、减速、或者停止)什么稳定状态。在这个过程中，也可以考虑其它因素来确定自动驾驶车辆100的速度，诸如，自动驾驶车辆100在行驶的道路中的横向位置、道路的曲率、静态和动态目标的接近度等等。Alternatively, the autonomous vehicle 100 or a computing device associated with the autonomous vehicle 100 (such as the computer system 160, computer vision system 132, memory 163 of FIG. For example, traffic, rain, ice on the road, etc.) to predict the behavior of the identified objects. Optionally, each identified object is dependent on the behavior of the other, so all identified objects can also be considered together to predict the behavior of a single identified object. The autonomous vehicle 100 is able to adjust its speed based on the predicted behavior of the identified object. In other words, the autonomous vehicle is able to determine what steady state the autonomous vehicle will need to adjust to (eg, accelerate, decelerate, or stop) based on the predicted behavior of the target. During this process, other factors may also be considered to determine the speed of the autonomous vehicle 100 , such as the lateral position of the autonomous vehicle 100 on the driving road, the curvature of the road, the proximity of static and dynamic objects, and so on.

除了提供调整自动驾驶汽车的速度的指令之外，计算设备还可以提供修改自动驾驶车辆100的转向角的指令，以使得自动驾驶汽车遵循给定的轨迹和/或维持与自动驾驶汽车附近的目标(例如，道路上的相邻车道中的轿车)的安全横向和纵向距离。In addition to providing instructions to adjust the speed of the self-driving car, the computing device may also provide instructions to modify the steering angle of the self-driving vehicle 100 so that the self-driving car follows a given trajectory and/or maintains a target near the self-driving car. (for example, cars in adjacent lanes on the road) safe lateral and longitudinal distances.

上述自动驾驶车辆100可以为轿车、卡车、摩托车、公共汽车、船、飞机、直升飞机、割草机、娱乐车、游乐场自动驾驶车辆、施工设备、电车、高尔夫球车、火车、和手推车等，本申请实施例不做特别的限定。The above-mentioned self-driving vehicle 100 may be a car, truck, motorcycle, bus, boat, airplane, helicopter, lawn mower, recreational vehicle, amusement park self-driving vehicle, construction equipment, tram, golf cart, train, and The trolley and the like are not particularly limited in this embodiment of the present application.

本申请实施例提供的目标车辆转向意图的确定方法，其中目标车辆可以是直排轮车，以下对直排轮车的概念进行介绍。In the method for determining a steering intention of a target vehicle provided in an embodiment of the present application, the target vehicle may be a straight-wheel vehicle, and the concept of a straight-wheel vehicle will be introduced below.

直排轮车包括至少两个车轮，该至少两个车轮处于同一排，或者说在前车轮转向角度为零时该至少两个车轮处于同一个平面。该至少两个车轮包括一个前车轮、以及一个或多个后车轮。后车轮可以与车架固定在一起，可随车架摆动。前车轮可以与车把固定在一起，可随车把摆动。The straight-wheel vehicle includes at least two wheels, and the at least two wheels are in the same row, or in other words, the at least two wheels are in the same plane when the steering angle of the front wheels is zero. The at least two wheels include a front wheel, and one or more rear wheels. The rear wheel can be fixed together with the vehicle frame and can swing with the vehicle frame. The front wheel can be fixed with the handlebar and can swing with the handlebar.

直排轮车可以是人力车、电动车或燃油车。直排轮车可以包括多个种类。以下结合附图对直排轮车的几种类型进行举例说明。例如，如图2a所示，直排轮车可以为两轮自行车。直排轮车还可以为三轮自行车，包括一个前车轮和两个后车轮，且三个车轮在同一排。如图2b所示，直排轮车可以为儿童平衡车。如图2c所示，直排轮车还可以为两轮滑板车。另外，直排轮车还可以为摩托车或电动自行车等。Inline-wheeled vehicles can be rickshaws, electric vehicles or gasoline vehicles. Inline-wheeled vehicles can include many types. Several types of inline-wheeled vehicles are illustrated below in conjunction with the accompanying drawings. For example, as shown in Figure 2a, the inline wheeled vehicle may be a two-wheeled bicycle. The inline-wheeled vehicle can also be a three-wheeled bicycle, comprising a front wheel and two rear wheels, and the three wheels are in the same row. As shown in Figure 2b, the inline wheeled vehicle can be a children's balance vehicle. As shown in Fig. 2c, the inline-wheeled vehicle can also be a two-wheeled scooter. In addition, the straight-wheel vehicle can also be a motorcycle or an electric bicycle.

本申请实施例中，以直排轮车为自行车为例进行说明，可以理解所述方案可以应用到任意直排轮车中。In the embodiment of the present application, an inline-wheeled vehicle is taken as an example for description, and it can be understood that the solution can be applied to any inline-wheeled vehicle.

如图3所示，为本申请提供的一种可能的应用场景示意图。目标车辆的转向会直接影响自动驾驶车辆的安全行驶。自动驾驶车辆可以根据目标车辆的转向意图确定行驶策略，以达到避障的目的。假设自动驾驶车辆错误判断目标车辆的转向意图，则自动驾驶车辆会制定错误的自动驾驶策略，影响正常行驶甚至导致交通事故。假设自动驾驶车辆无法判断目标车辆的转向意图，也会无法制定合理的驾驶策略，影响驾驶的安全性。当目标车辆进入自动驾驶车辆的视觉范围内时，例如，目标车辆进入图3中虚线夹角标注的范围内时，自动驾驶车辆获取目标车辆的视觉信息，通过视觉信息判断目标车辆的转向意图，以制定安全合理的驾驶策略。As shown in FIG. 3 , it is a schematic diagram of a possible application scenario provided by this application. The steering of the target vehicle will directly affect the safe driving of the autonomous vehicle. The autonomous vehicle can determine the driving strategy according to the steering intention of the target vehicle to achieve the purpose of obstacle avoidance. Assuming that the self-driving vehicle misjudges the steering intention of the target vehicle, the self-driving vehicle will formulate a wrong self-driving strategy, affecting normal driving and even causing traffic accidents. Assuming that the self-driving vehicle cannot judge the steering intention of the target vehicle, it will also be unable to formulate a reasonable driving strategy, which will affect the safety of driving. When the target vehicle enters the visual range of the self-driving vehicle, for example, when the target vehicle enters the range marked by the angle between the dotted lines in Figure 3, the self-driving vehicle obtains the visual information of the target vehicle, and judges the steering intention of the target vehicle through the visual information. To formulate a safe and reasonable driving strategy.

如图4所示，下面对本申请实施例提供的目标车辆转向意图的确定方法的具体流程进行说明。该方法可以由自动驾驶车辆执行。该方法也可以由自动驾驶车辆控制装置、电子设备或车载设备执行，或者，该方法可以由自动驾驶车辆的部件或者与自动驾驶车辆相关的设备执行，执行该方法的设备可以包括执行自动驾驶算法的芯片，例如人工智能(artificial intelligence，AI)芯片、图形处理器(graphics processing unit，GPU)芯片、中央处理器(central processing unit，CPU)等芯片，也可以是包括其中多种芯片构成的***。As shown in FIG. 4 , the specific flow of the method for determining the steering intention of the target vehicle provided by the embodiment of the present application will be described below. The method can be performed by an autonomous vehicle. The method may also be executed by a self-driving vehicle control device, an electronic device, or an on-board device, or the method may be executed by a component of a self-driving vehicle or a device related to a self-driving vehicle, and the device performing the method may include an automatic driving algorithm Chips, such as artificial intelligence (AI) chips, graphics processing unit (GPU) chips, central processing unit (central processing unit, CPU) and other chips, can also be a system composed of multiple chips. .

S401、获取目标车辆的视觉信息。S401. Obtain visual information of a target vehicle.

视觉信息包括视觉图像或者视频等。例如，可以通过车辆中的相机传感器获取目标车辆的视觉信息。可选的，该车辆可以为自动驾驶车辆，或者集成了ADAS的车辆。Visual information includes visual images or videos, and the like. For example, the visual information of the target vehicle can be acquired through the camera sensor in the vehicle. Optionally, the vehicle may be an autonomous driving vehicle, or a vehicle integrated with ADAS.

S402、根据视觉信息中目标车辆的姿态信息，姿态信息可以是表示目标车辆的车身姿态的信息。S402. According to the attitude information of the target vehicle in the visual information, the attitude information may be information representing the body attitude of the target vehicle.

目标车辆的姿态信息能够体现目标车辆在空间的运动姿态。The attitude information of the target vehicle can reflect the movement attitude of the target vehicle in space.

S403、根据目标车辆的姿态信息，确定目标车辆的转向意图。S403. Determine the steering intention of the target vehicle according to the attitude information of the target vehicle.

可选的，在S403之后，还包括S404。Optionally, after S403, S404 is also included.

S404、根据所述目标车辆的转向意图，确定安全驾驶策略。安全驾驶策略也可以称为自动驾驶策略，可以用于自动驾驶车辆预估风险，保证自动驾驶车辆在道路上行驶的安全性。示例地，执行该转向意图确定方法的终端，可以相应的采取避开目标车辆的策略，如控制该终端停止避开目标车辆，或者控制该终端转向避开目标车辆等。S404. Determine a safe driving strategy according to the steering intention of the target vehicle. The safe driving strategy can also be called the automatic driving strategy, which can be used to estimate the risk of the self-driving vehicle and ensure the safety of the self-driving vehicle on the road. For example, the terminal executing the method for determining the turning intention may adopt a corresponding strategy to avoid the target vehicle, such as controlling the terminal to stop avoiding the target vehicle, or controlling the terminal to turn to avoid the target vehicle.

目标车辆的姿态信息能够更精准的表达目标车辆的转向意图，这样，所确定的目标车辆的转向意图能够更加精确，有助于提高预测目标车辆的转向意图的精确度，更好的辅助自动驾驶，保证自动驾驶车辆在道路上的行驶安全性。The attitude information of the target vehicle can express the steering intention of the target vehicle more accurately. In this way, the determined steering intention of the target vehicle can be more accurate, which helps to improve the accuracy of predicting the steering intention of the target vehicle and better assist automatic driving. , to ensure the driving safety of self-driving vehicles on the road.

应理解，上述方法可在仅有单帧图像的情况下确定目标车辆的转向意图，但在实际应用中由于检测精度、外部扰动和***噪声的影响，还可以通过多帧图像用本申请所述的方法进行处理，如，分别对多帧图像的部分帧图像或者全部帧图像，根据本申请提供的转向意图确定方法，对每一帧的图像分别处理得到每帧图像对应的目标车辆的姿态信息，再根据各个姿态信息进行筛选，或者融合处理等确定目标车辆最终的姿态信息，示例地利用卡尔曼滤波器对目标车辆的姿态信息(侧倾角，转向角)进行滤波，其中卡尔曼滤波器通常假设转向角恒定或转向角速度恒定。或者，先对多帧图像进行滤波(筛选)，找到合适一帧或者部分帧图像，再对合适一帧图像或者部分帧图像应用本申请所述的方法。It should be understood that the above method can determine the steering intention of the target vehicle in the case of only a single frame image, but in practical applications, due to the influence of detection accuracy, external disturbance and system noise, it can also be used through multiple frames of images using the method described in this application. For example, for partial frame images or all frame images of multi-frame images, according to the steering intention determination method provided by this application, each frame image is processed separately to obtain the attitude information of the target vehicle corresponding to each frame image , and then filter according to each attitude information, or determine the final attitude information of the target vehicle through fusion processing, for example, use a Kalman filter to filter the attitude information (roll angle, steering angle) of the target vehicle, wherein the Kalman filter is usually Assume constant steering angle or constant steering angular velocity. Or, filter (screen) multiple frames of images first to find a suitable frame or part of the frame images, and then apply the method described in this application to the suitable frame of images or part of the frame images.

以下对图4实施例的一些可能的实现方式进行说明。Some possible implementations of the embodiment in FIG. 4 are described below.

视觉信息可以是包含目标车辆的视觉图像，该视觉信息可以是自动驾驶车辆上的摄像头获取的，例如也可以是图1所示的自动驾驶车辆100中的相机传感器123获取的。该视觉信息是二维的图像信息。The visual information may be a visual image containing the target vehicle, and the visual information may be acquired by a camera on the autonomous vehicle, for example, it may also be acquired by the camera sensor 123 in the autonomous vehicle 100 shown in FIG. 1 . This visual information is two-dimensional image information.

视觉信息可以是对自动驾驶车辆获取的图像截取的边界框(bounding box)。自动驾驶车辆在获取范围内可获取原图像，可能包括除目标车辆之外的其余事物的图像。在自动驾驶车辆获取的原图像中，截取边界框。以目标车辆为目标直排轮车为例，如图5所示，该边界框包含目标直排轮车，或者该边界框包含驾驶该目标直排轮车的人以及该目标直排轮车。可以理解为，边界框为自动驾驶车辆获取的原图像中包括目标直排轮车的小图像，这样，通过可以检测边界框中目标直排轮车的姿态信息，可以排除除目标直排轮车之外的其余事物的干扰，更利于检测目标直排轮车的姿态信息。The visual information can be a bounding box of an image capture captured by an autonomous vehicle. An autonomous vehicle acquires raw images within range, which may include images of other things in addition to the target vehicle. In the original image acquired by the self-driving vehicle, the bounding box is intercepted. Taking the target vehicle as the target inline-wheeled vehicle as an example, as shown in FIG. 5 , the bounding box includes the target inline-wheeled vehicle, or the bounding box includes the person driving the target inline-wheeled vehicle and the target inline-wheeled vehicle. It can be understood that the bounding box is a small image of the target inline-wheeled vehicle in the original image acquired by the self-driving vehicle. In this way, by detecting the attitude information of the target inline-wheeled vehicle in the bounding box, it is possible to exclude the target inline-wheeled vehicle. The interference of other things is more conducive to detecting the attitude information of the target inline-wheeled vehicle.

以下对S402中根据视觉信息中确定目标车辆的姿态信息的可能实现方式进行说明。A possible implementation manner of determining the attitude information of the target vehicle according to the visual information in S402 will be described below.

目标车辆的姿态信息可以由目标车辆在视觉信息中的几何形状确定。具体地，可以确定视觉信息中目标车辆的几何形状，根据几何形状，确定目标车辆的姿态信息。示例地，视觉信息为二维图像时，目标车辆在二维图像中对应的几何形状可以是目标车辆的车轮在视觉信息中对应的第一几何形状。例如，目标车辆为直排轮车，目标直排轮车的车轮在二维图像中对应的第一几何形状为椭圆。在一种可能的实现方式中，目标直排轮车的姿态信息可以由目标直排轮车在视觉信息中的车身参数确定，车身参数可以是目标直排轮车的车轮在二维图像坐标系中对应的椭圆的参数。椭圆在二维图像坐标系中的参数可以包括长半轴a、短半轴b、中心位置(x ₀，y ₀)以及方位角ρ。 The pose information of the target vehicle can be determined by the geometric shape of the target vehicle in the visual information. Specifically, the geometric shape of the target vehicle in the visual information can be determined, and the attitude information of the target vehicle can be determined according to the geometric shape. For example, when the visual information is a two-dimensional image, the geometric shape corresponding to the target vehicle in the two-dimensional image may be the first geometric shape corresponding to the wheel of the target vehicle in the visual information. For example, the target vehicle is an inline vehicle, and the first geometric shape corresponding to the wheels of the target inline vehicle in the two-dimensional image is an ellipse. In a possible implementation, the attitude information of the target inline-wheeled vehicle can be determined by the body parameters of the target inline-wheeled vehicle in the visual information, and the body parameters can be the wheels of the target inline-wheeled vehicle in the two-dimensional image coordinate system The parameters of the corresponding ellipse in . The parameters of the ellipse in the two-dimensional image coordinate system may include semi-major axis a, semi-minor axis b, center position (x ₀ , y ₀ ) and azimuth ρ.

目标车辆的姿态信息是三维空间坐标系的信息，根据目标车辆的在视觉信息中对应的几何形状确定目标车辆的姿态信息，可以通过将二维图像坐标系中的信息转变到三维空间坐标系的信息实现。The attitude information of the target vehicle is the information of the three-dimensional space coordinate system. According to the geometric shape corresponding to the target vehicle in the visual information, the attitude information of the target vehicle can be determined by converting the information in the two-dimensional image coordinate system to the three-dimensional space coordinate system. information realization.

姿态信息包括侧倾角。侧倾角是用于描述姿态信息的一种参数，目标车辆的侧倾角与目标车辆的横向加速度有直接联系，目标车辆的横向加速度能够表征目标车辆的转向意图，从而可知，根据目标车辆的侧倾角能够确定目标车辆的转向意图。相比于传统的目标车辆轨迹预测方法，根据目标车辆的侧倾角确定目标车辆的转向意图，无需任何历史轨迹的获取。轨迹预测方法需要机器学习，机器学习的方法对于训练数据和道路环境的要求严格，本申请实施例通过目标车辆的侧倾角确定目标车辆的转向意图，利用视觉图像确定车身姿态以对目标车辆转向意图进行识别，适用范围更广，提高预测精度，以更好的提高自动驾驶车辆的安全驾驶性能。并且，该方法无需历史轨迹的学习，能够提高预测效率。Attitude information includes roll angle. The roll angle is a parameter used to describe the attitude information. The roll angle of the target vehicle is directly related to the lateral acceleration of the target vehicle. The lateral acceleration of the target vehicle can represent the steering intention of the target vehicle. The steering intent of the target vehicle can be determined. Compared with the traditional target vehicle trajectory prediction method, the steering intention of the target vehicle is determined according to the roll angle of the target vehicle without any historical trajectory acquisition. The trajectory prediction method requires machine learning. The machine learning method has strict requirements on the training data and the road environment. In the embodiment of the present application, the steering intention of the target vehicle is determined by the roll angle of the target vehicle, and the body posture is determined by using the visual image to predict the steering intention of the target vehicle. Recognition, a wider range of applications, improved prediction accuracy, in order to better improve the safe driving performance of self-driving vehicles. Moreover, this method does not require the learning of historical trajectories, which can improve the prediction efficiency.

根据目标车辆的在视觉信息中对应的几何形状确定目标车辆的姿态信息的方法，以下介绍两种实现方式，包括实现方式A和实现方式B。The method for determining the attitude information of the target vehicle according to the geometric shape corresponding to the target vehicle in the visual information, two implementations are introduced below, including implementation A and implementation B.

在一种实现方式A中，目标车辆的在视觉信息中对应的几何形状为二维图像坐标系中的信息，通过将二维图像坐标系中的信息转变到三维空间坐标系的信息的方法，将目标车辆的在视觉信息中对应的几何形状转变为目标车辆所在的平面，记为第一平面。根据第一平面确定目标车辆的姿态信息。第一平面可以是目标车辆的后车轮所在的平面，也可以是目标车辆的车架所在的平面。目标车辆的后车轮与车架在同一个平面。假设垂直于地面的平面为第二平面，则第一平面与第二平面之间的夹角可以认为是侧倾角。侧倾角可以用于判断目标车辆的转向意图。In one implementation A, the geometric shape corresponding to the visual information of the target vehicle is the information in the two-dimensional image coordinate system, and by converting the information in the two-dimensional image coordinate system to the information in the three-dimensional space coordinate system, The geometric shape corresponding to the target vehicle in the visual information is transformed into the plane where the target vehicle is located, which is recorded as the first plane. The attitude information of the target vehicle is determined according to the first plane. The first plane may be the plane where the rear wheels of the target vehicle are located, or the plane where the frame of the target vehicle is located. The rear wheels of the target vehicle are in the same plane as the frame. Assuming that the plane perpendicular to the ground is the second plane, the angle between the first plane and the second plane can be considered as the roll angle. The roll angle can be used to judge the steering intention of the target vehicle.

在一种实现方式B中，假设目标车辆的车轮在二维图像中显示为椭圆，检测视觉信息中目标车辆的车轮在二维图像坐标系对应的椭圆，根据椭圆确定目标车辆的车轮在三维相机坐标系中所在平面的法向量，根据目标车辆的车轮在三维相机坐标系中所在平面的法向量确定目标车辆的姿态信息，目标车辆的车轮在三维相机坐标系中所在平面即第一平面。In one implementation B, assuming that the wheels of the target vehicle are displayed as ellipses in the two-dimensional image, the ellipse corresponding to the wheels of the target vehicle in the two-dimensional image coordinate system in the visual information is detected, and the wheels of the target vehicle are determined according to the ellipse in the three-dimensional camera The normal vector of the plane in the coordinate system determines the attitude information of the target vehicle according to the normal vector of the plane where the wheels of the target vehicle are located in the three-dimensional camera coordinate system. The plane where the wheels of the target vehicle are located in the three-dimensional camera coordinate system is the first plane.

目标车辆的车轮在三维相机坐标系中所在平面的法向量，可以简述为目标车辆的车轮的空间法向量。目标车辆的后车轮所在的平面即目标车辆的车身或车架所在的平面，目标车辆的后车轮的空间法向量，即目标车辆的车身的空间法向量，或者目标车辆的车架的空间法向量，或者简述为目标车辆的空间法向量。The normal vector of the plane where the wheel of the target vehicle is located in the three-dimensional camera coordinate system can be briefly described as the space normal vector of the wheel of the target vehicle. The plane where the rear wheels of the target vehicle are located is the plane where the body or frame of the target vehicle is located, and the space normal vector of the rear wheels of the target vehicle is the space normal vector of the body of the target vehicle, or the space normal vector of the vehicle frame of the target vehicle , or briefly referred to as the space normal vector of the target vehicle.

如图6所示，以下举例说明二维图像坐标系中的二维椭圆转换到三维相机坐标系的信息。二维图像坐标系的原点为o，横轴为v，纵轴为u。三维相机坐标系的原点为o ^c，三轴分别用x ^c、y ^c和z ^c表示。将二维图像中的椭圆转换到三维相机坐标系中，可以通过以下方式实现。利用相机投影模型将椭圆转换为三维相机坐标系中的椭圆锥面。椭圆锥面可以表达为：Ax ²+By ²+Cxy+Dxz+Eyz+Fz ²＝0。其中，A、B、C、D、E、F为椭圆锥面的几何参数，x、y、z为椭圆锥面对应的三轴的坐标位置。可选地，参数A、B、C、D、E、F是根据二维图像坐标系中的椭圆参数确定，椭圆参数可以包括长半轴a、短半轴b、中心位置(x ₀，y ₀)以及方位角ρ中的一种或者多种。利用空间圆姿态测量方法从椭圆锥面中确定第一平面(图6中示意为车轮)，第一平面与椭圆锥面的交线为正空间圆，第一平面的法向量即目标车辆的车轮在三维相机坐标系中所在平面的法向量。 As shown in FIG. 6 , the following example illustrates the transformation of a two-dimensional ellipse in a two-dimensional image coordinate system into a three-dimensional camera coordinate system. The origin of the two-dimensional image coordinate system is o, the horizontal axis is v, and the vertical axis is u. The origin of the 3D camera coordinate system is o ^c , and the three axes are denoted by x ^c , y ^c and z ^c respectively. Converting the ellipse in the 2D image to the 3D camera coordinate system can be achieved in the following way. The ellipse is transformed into an elliptical cone in the 3D camera coordinate system using the camera projection model. The elliptical cone can be expressed as: Ax ² +By ² +Cxy+Dxz+Eyz+Fz ² =0. Among them, A, B, C, D, E, and F are the geometric parameters of the elliptical cone, and x, y, and z are the coordinate positions of the three axes corresponding to the ellipse. Optionally, the parameters A, B, C, D, E, and F are determined according to the ellipse parameters in the two-dimensional image coordinate system, and the ellipse parameters may include the semi-major axis a, the semi-minor axis b, the center position (x ₀ , y ₀ ) and one or more of the azimuth ρ. Utilize the spatial circular attitude measurement method to determine the first plane (shown as a wheel in Figure 6) from the elliptical cone surface, the intersection line between the first plane and the elliptical cone surface is a positive space circle, and the normal vector of the first plane is the wheel of the target vehicle The normal vector of the plane in the 3D camera coordinate system.

其中，椭圆锥面存在多个截面，截面与椭圆锥面的交线可能为椭圆或正空间圆，与椭圆锥面的交线为正空间圆的平面也可能存在多个。任意一个与椭圆锥面的交线为正空间圆的平面所对应的姿态信息都是一样的，所确定出来的倾向角都是相同的。Wherein, there are multiple cross-sections of the elliptical conical surface, the intersection line of the cross-section and the elliptical conical surface may be an ellipse or a positive space circle, and there may also be multiple planes whose intersection lines with the elliptical conical surface are a positive space circle. The attitude information corresponding to any plane whose intersection line with the elliptical cone surface is a positive space circle is the same, and the determined inclination angles are all the same.

三维空间坐标系也可以称为三维相机坐标系。当自动驾驶车辆的摄像机外部参数确定后，三维相机坐标系和二维图像坐标系的关系就确定了。三维空间坐标系中的目标车辆的车轮，在光的反射中被记录下来，反射的光在二维图像上映射出来。相机投影模型能够将二维的几何形状转换为三维的几何形状。例如，二维图像上的信息为椭圆，相机投影模型可以将椭圆转换为三维的椭圆锥面。The three-dimensional space coordinate system may also be called a three-dimensional camera coordinate system. When the external parameters of the camera of the self-driving vehicle are determined, the relationship between the three-dimensional camera coordinate system and the two-dimensional image coordinate system is determined. The wheels of the target vehicle in the three-dimensional space coordinate system are recorded in the reflection of light, and the reflected light is mapped on the two-dimensional image. The camera projection model is capable of transforming two-dimensional geometric shapes into three-dimensional geometric shapes. For example, the information on a two-dimensional image is an ellipse, and the camera projection model can convert the ellipse into a three-dimensional elliptical cone.

由此，目标车辆的前车轮和后车轮的空间法向量，都可以根据上述方法确定。Thus, the space normal vectors of the front wheels and rear wheels of the target vehicle can be determined according to the above method.

以下对确定目标车辆在视觉信息中对应的几何形状的可能实现方式进行说明。视觉信息为二维图像信息，目标车辆的车轮在二维图像中显示为椭圆。检测视觉信息中的目标车辆的车轮对应的椭圆，视觉信息可以是自动驾驶车辆获取的边界框(bounding box)。其中，视觉信息中可能存在多个椭圆，非车轮对应的椭圆可以认为是椭圆噪声。去除椭圆噪声有助于提高检测效果。本申请实施例中，可以通过以下方法检测视觉信息中目标车辆的车轮对应的椭圆。可以根据预定的算法提取出视觉信息中的椭圆，例如，采用弧支持椭圆检测算法提取视觉信息中的椭圆。二维图像中的椭圆可以根据以下5个参数确定：长半轴a、短半轴b、中心位置(x ₀，y ₀)以及方位角ρ。在视觉信息中初始检测出来多个椭圆，假设多个椭圆的数量为n，n为大于等于2的整数，多个椭圆称为椭圆集合。椭圆集合可以用e _init,i表示，e _init,i＝(a,b,(x ₀，y ₀)，ρ)，i＝1,2，…，n。可以按照约束条件在椭圆集合中确定目标车辆的车轮在视觉信息中对应的第一椭圆，约定条件包括以下一项或多项： A possible implementation manner of determining the geometric shape corresponding to the target vehicle in the visual information will be described below. The visual information is two-dimensional image information, and the wheels of the target vehicle are displayed as ellipses in the two-dimensional image. The ellipse corresponding to the wheel of the target vehicle in the visual information is detected, and the visual information may be a bounding box (bounding box) acquired by the self-driving vehicle. Among them, there may be multiple ellipses in the visual information, and the ellipses not corresponding to the wheels can be considered as elliptical noise. Removing elliptic noise helps to improve the detection performance. In the embodiment of the present application, the following method can be used to detect the ellipse corresponding to the wheel of the target vehicle in the visual information. The ellipse in the visual information can be extracted according to a predetermined algorithm, for example, an arc-supported ellipse detection algorithm is used to extract the ellipse in the visual information. The ellipse in the two-dimensional image can be determined according to the following five parameters: semi-major axis a, semi-minor axis b, center position (x ₀ , y ₀ ) and azimuth ρ. Multiple ellipses are initially detected in the visual information. Assume that the number of multiple ellipses is n, and n is an integer greater than or equal to 2. The multiple ellipses are called an ellipse set. The set of ellipses can be represented by e _init,i , e _init,i =(a,b,(x ₀ ,y ₀ ), ρ), i=1,2,...,n. The first ellipse corresponding to the wheel of the target vehicle in the visual information can be determined in the ellipse set according to the constraints, and the agreed conditions include one or more of the following:

1)椭圆在视觉信息中的尺寸占比满足设定的比例范围。1) The size ratio of the ellipse in the visual information meets the set ratio range.

假设视觉信息为边界框，车轮对应的椭圆的大小在边界框中的高度占比具有一定合理的比例，这样可以去除一些过大或过小的椭圆。该设定的比例范围例如可以是1/3、或2/5，或其他合理的比例。Assuming that the visual information is a bounding box, the size of the ellipse corresponding to the wheel has a certain reasonable ratio to the height of the bounding box, so that some ellipses that are too large or too small can be removed. The set ratio range may be, for example, 1/3, or 2/5, or other reasonable ratios.

2)椭圆的中心位置在视觉信息的中心线的下方，中心线将视觉信息分为上下两部分。可选的，所述中心线将所述视觉信息分为均匀的上下两部分。2) The center position of the ellipse is below the center line of the visual information, and the center line divides the visual information into upper and lower parts. Optionally, the central line divides the visual information into even upper and lower parts.

目标直排轮车的前车轮会在边界框的左下方，后车轮会在边界框的右下方，因此，将椭圆的中心位置约束在中心线的下方，能够去除一些由背景造成的椭圆噪声。The front wheel of the target inline vehicle will be at the bottom left of the bounding box, and the rear wheel will be at the bottom right of the bounding box, so constraining the center position of the ellipse to be below the centerline removes some ellipse noise caused by the background.

3)椭圆的边缘的颜色趋势符合：边缘外侧为深，边缘内侧为浅。3) The color trend of the edge of the ellipse conforms to: the outer edge is darker, and the inner edge is lighter.

目标车辆的车轮在二维图像中对应的椭圆，可能包括轮胎对应的椭圆、挡泥板对应的椭圆、或轮胎与轮毂交界的边沿对应的椭圆。而轮胎与轮毂交界的边沿对应的椭圆的参数更加精确，更能反映目标车辆的转向意图。在视觉信息中轮胎与轮毂交界的边沿的颜色趋势符合“外深内浅”，即边缘外侧为深，边缘内侧为浅。通过“外深内浅”的极性约束，能够选择轮胎与轮毂交界的边沿对应的椭圆。也即是，目标车辆的车轮在视觉信息中对应的椭圆，该椭圆的边缘的颜色符合：边缘外侧的颜色深于边缘内侧的颜色。The ellipse corresponding to the wheel of the target vehicle in the two-dimensional image may include an ellipse corresponding to the tire, an ellipse corresponding to the fender, or an ellipse corresponding to the border between the tire and the hub. The parameters of the ellipse corresponding to the border of the tire and the wheel hub are more accurate and can better reflect the steering intention of the target vehicle. In the visual information, the color trend of the edge of the junction between the tire and the hub conforms to "dark on the outside and light on the inside", that is, the outside of the edge is dark, and the inside of the edge is light. Through the polarity constraint of "deep on the outside and shallow on the inside", the ellipse corresponding to the border between the tire and the hub can be selected. That is, for the ellipse corresponding to the wheel of the target vehicle in the visual information, the color of the edge of the ellipse conforms to: the color outside the edge is darker than the color inside the edge.

基于图5的边界框的举例，如图7所示，在图7的(a)中显示了轮胎对应的椭圆、挡泥板对应的椭圆、或轮胎与轮毂交界的边沿对应的椭圆，通过上述3)的约束，最终得到图7的(b)显示的椭圆。Based on the example of the bounding box in Fig. 5, as shown in Fig. 7, the ellipse corresponding to the tire, the ellipse corresponding to the fender, or the ellipse corresponding to the border between the tire and the hub is shown in (a) of Fig. 7, through the above 3) constraints, the ellipse shown in (b) of Figure 7 is finally obtained.

以下给出一种姿态信息的计算方法。A calculation method of attitude information is given below.

姿态信息包括侧倾角，如图8所示，侧倾角用

表示，

为目标车辆的后车轮所在第一平面与垂直地面的第二平面之间的夹角，第一平面与第二平面的交线为第一平面与地面的交线。 The attitude information includes the roll angle, as shown in Figure 8, the roll angle is used

express,

is the angle between the first plane where the rear wheels of the target vehicle are located and the second plane perpendicular to the ground, and the intersection line between the first plane and the second plane is the intersection line between the first plane and the ground.

侧倾角可以通过目标车辆的后车轮的空间法向量以及地面法向量确定。假设目标车辆的后车轮的空间法向量用n _r表示，地面法向量用n _g表示，那么侧倾角

符合以下公式(1)： The roll angle can be determined from the space normal vectors of the rear wheels of the target vehicle and the ground normal vectors. Assuming that the space normal vector of the rear wheel of the target vehicle is represented by n _r , and the ground normal vector is represented by n _g , then the roll angle

According to the following formula (1):

姿态信息还可以包括转向角。转向角为目标车辆的前车轮的偏转角度。目标车辆的前车轮所在平面为第三平面，目标车辆的后车轮所在的平面为第一平面，转向角可以是第三平面与第一平面之间的夹角。Attitude information may also include steering angle. The steering angle is the deflection angle of the front wheels of the target vehicle. The plane where the front wheels of the target vehicle are located is the third plane, the plane where the rear wheels of the target vehicle is located is the first plane, and the steering angle may be the angle between the third plane and the first plane.

如图8所示，转向角用δ表示，δ为第三平面与第一平面之间的夹角。As shown in FIG. 8 , the steering angle is represented by δ, and δ is the angle between the third plane and the first plane.

转向角可以通过目标车辆的前车轮的空间法向量和后车轮的空间法向量确定。例如，转向角为目标车辆的前车轮的空间法向量和后车轮的空间法向量的夹角。假设目标车辆的后车轮的空间法向量用n _r表示，前车轮的空间法向量用n _f表示，转向角δ符合下述公式(2)： The steering angle can be determined by the space normal vectors of the front wheels and the space normal vectors of the rear wheels of the target vehicle. For example, the steering angle is the included angle between the space normal vectors of the front wheels and the space normal vectors of the rear wheels of the target vehicle. Assuming that the space normal vector of the rear wheel of the target vehicle is denoted by _nr , and the space normal vector of the front wheel is denoted by _nf , the steering angle δ conforms to the following formula (2):

姿态信息还可以包括目标车辆的车身长度，目标车辆的车身长度可以是目标车辆的轴距。轴距可以为目标车辆的前车轮和后车轮的空间距离。轴距可以在视觉信息中检测，也可以预先设定。轴距与目标车辆的车型有关，轴距可以是预设的固定值，或者经验值。例如，自行车的轴距可以设置为1米2，或1米5。The attitude information may also include the body length of the target vehicle, and the body length of the target vehicle may be the wheelbase of the target vehicle. The wheelbase may be the spatial distance between the front wheels and the rear wheels of the target vehicle. The wheelbase can be detected in the visual information or can be preset. The wheelbase is related to the model of the target vehicle, and the wheelbase can be a preset fixed value or an empirical value. For example, the wheelbase of a bicycle can be set to 1 meter 2 , or 1 meter 5 .

转向角、轴距或两者的结合，可以用于确定目标车辆的转弯半径，从而自动驾驶车辆可以根据目标车辆的转弯半径，确定安全驾驶策略。Steering angle, wheelbase, or a combination of the two can be used to determine the turning radius of the target vehicle, so that the autonomous vehicle can determine a safe driving strategy based on the turning radius of the target vehicle.

以下对根据侧倾角确定目标车辆的转向意图的可能实现方式。目标车辆后车轮所在的平面为第一平面，垂直地面的平面为第二平面The following is a possible implementation of determining the steering intention of the target vehicle according to the roll angle. The plane where the rear wheels of the target vehicle are located is the first plane, and the plane perpendicular to the ground is the second plane

在一个可能的设计中，以第二平面为参考的平面，当目标车辆位于第二平面时，侧倾角为零。以前车轮的方向为基准，第二平面向右为正，向左为负。侧倾角的范围为-90°至90°。In a possible design, taking the second plane as a reference plane, when the target vehicle is located on the second plane, the roll angle is zero. Based on the direction of the front wheel, the second plane is positive to the right and negative to the left. The roll angle ranges from -90° to 90°.

可以设定第一阈值和第二阈值，第一阈值为正，第二阈值为负。A first threshold and a second threshold can be set, the first threshold is positive, and the second threshold is negative.

基于图9a的示例：Example based on Figure 9a:

当侧倾角小于第一阈值时，确定目标车辆的转向意图为左转。这种情况下，目标车辆的车身姿态为左侧倾，视觉信息如图10的(a)所示。When the roll angle is smaller than the first threshold, it is determined that the steering intention of the target vehicle is to turn left. In this case, the body posture of the target vehicle is tilted to the left, and the visual information is shown in (a) of Figure 10 .

当侧倾角大于第一阈值小于第二阈值时，确定目标车辆的转向意图为直行。这种情况下，目标车辆的车身姿态为竖直，视觉信息如图10的(b)所示。When the roll angle is larger than the first threshold and smaller than the second threshold, it is determined that the steering intention of the target vehicle is to go straight. In this case, the body posture of the target vehicle is vertical, and the visual information is shown in (b) of FIG. 10 .

当侧倾角大于第二阈值时，确定目标车辆的转向意图为右转。这种情况下，目标车辆的车身姿态为右侧倾，视觉信息如图10的(c)所示。When the roll angle is greater than the second threshold, it is determined that the steering intention of the target vehicle is a right turn. In this case, the body posture of the target vehicle is tilted to the right, and the visual information is shown in (c) of FIG. 10 .

示例地，第一阈值为-5°，第二阈值为6°，当侧倾角小于-5°时，确定目标车辆的转向意图为左转，当侧倾角大于等于-5°且小于等于6°时，确定目标车辆的转向意图为直行，当侧倾角大于6°时，确定目标车辆的转向意图为右转。可选的，当侧倾角等于-5°时，确定目标车辆的转向意图为左转，或者，当侧倾角等于6°时，确定目标车辆的转向意图为右转。For example, the first threshold is -5°, and the second threshold is 6°. When the roll angle is less than -5°, it is determined that the steering intention of the target vehicle is a left turn. When the roll angle is greater than or equal to -5° and less than or equal to 6° , it is determined that the steering intention of the target vehicle is going straight, and when the roll angle is greater than 6°, it is determined that the steering intention of the target vehicle is a right turn. Optionally, when the roll angle is equal to -5°, it is determined that the target vehicle's steering intention is to turn left, or, when the roll angle is equal to 6°, it is determined that the target vehicle's steering intention is to turn right.

在另一个可能的设计中，地面为参考平面，当目标车辆位于第二平面时，侧倾角为90°，侧倾角的范围为0°至180°。第一阈值和第二阈值都为正，如图9b所示。In another possible design, the ground is a reference plane, and when the target vehicle is located on the second plane, the roll angle is 90°, and the roll angle ranges from 0° to 180°. Both the first threshold and the second threshold are positive, as shown in Figure 9b.

基于图9b的示例：Example based on Figure 9b:

情况一：第一阈值和第二阈值均表示为与90°之间的夹角，即第一阈值和第二阈值都可以表示为锐角。当侧倾角小于(90°-第一阈值)时，确定目标车辆的转向意图为左转。这种情况下，目标车辆的车身姿态为左侧倾，视觉信息如图10的(a)所示。Case 1: Both the first threshold and the second threshold are expressed as angles between 90°, that is, both the first threshold and the second threshold can be expressed as acute angles. When the roll angle is less than (90°-the first threshold), it is determined that the steering intention of the target vehicle is to turn left. In this case, the body posture of the target vehicle is tilted to the left, and the visual information is shown in (a) of Figure 10 .

当侧倾角介于(90°-第一阈值)与(90°+第二阈值)之间时，确定目标车辆的转向意图为直行。这种情况下，目标车辆的车身姿态为竖直或者近似竖直，视觉信息如图10的(b)所示。When the roll angle is between (90°-the first threshold) and (90°+the second threshold), it is determined that the steering intention of the target vehicle is to go straight. In this case, the body posture of the target vehicle is vertical or approximately vertical, and the visual information is shown in (b) of FIG. 10 .

当侧倾角大于(90°+第二阈值)时，确定目标车辆的转向意图为右转。这种情况下，目标车辆的车身姿态为右侧倾，视觉信息如图10的(c)所示。When the roll angle is greater than (90°+second threshold), it is determined that the steering intention of the target vehicle is a right turn. In this case, the body posture of the target vehicle is tilted to the right, and the visual information is shown in (c) of FIG. 10 .

示例地，第一阈值为12°，第二阈值为7°，当侧倾角小于等于(78°)时，确定目标车辆的转向意图为左转，当侧倾角大于78°且小于等于97°时，确定目标车辆的转向意图为直行，当侧倾角大于97°时，确定目标车辆的转向意图为右转。可选的，当侧倾角等于78°时，确定目标车辆的转向意图为直行，或者，当侧倾角等于97°时，确定目标车辆的转向意图为右转。For example, the first threshold is 12°, and the second threshold is 7°. When the roll angle is less than or equal to (78°), it is determined that the steering intention of the target vehicle is to turn left. When the roll angle is greater than 78° and less than or equal to 97° , it is determined that the steering intention of the target vehicle is going straight, and when the roll angle is greater than 97°, it is determined that the steering intention of the target vehicle is a right turn. Optionally, when the roll angle is equal to 78°, it is determined that the target vehicle's steering intention is to go straight, or, when the roll angle is equal to 97°, it is determined that the target vehicle's steering intention is to turn right.

情况二，第一阈值和第二阈值均表示为与0°之间的夹角。即第一阈值和第二阈值一个为锐角一个为钝角。当侧倾角小于第一阈值时，确定目标车辆的转向意图为左转；当侧倾角介于第一阈值和第二阈值之间时，确定目标车辆的转向意图为直行；当侧倾角大于第二阈值时，确定目标车辆的转向意图为右转。示例地，第一阈值为80°，第二阈值为100°，当侧倾角小于等于80°时，确定目标车辆的转向意图为左转，当侧倾角大于80°且小于等于100°时，确定目标车辆的转向意图为直行，当侧倾角大于100°时，确定目标车辆的转向意图为右转。可选的，当侧倾角等于80°时，确定目标车辆的转向意图为直行，或者，当侧倾角等于100°时，确定目标车辆的转向意图为右转。In the second case, both the first threshold and the second threshold are expressed as an included angle with 0°. That is, one of the first threshold and the second threshold is an acute angle and the other is an obtuse angle. When the roll angle is less than the first threshold, it is determined that the steering intention of the target vehicle is to turn left; when the roll angle is between the first threshold and the second threshold, it is determined that the steering intention of the target vehicle is going straight; When the threshold is reached, it is determined that the steering intention of the target vehicle is a right turn. For example, the first threshold is 80°, and the second threshold is 100°. When the roll angle is less than or equal to 80°, it is determined that the steering intention of the target vehicle is a left turn. When the roll angle is greater than 80° and less than or equal to 100°, it is determined that The steering intention of the target vehicle is to go straight, and when the roll angle is greater than 100°, it is determined that the steering intention of the target vehicle is to turn right. Optionally, when the roll angle is equal to 80°, it is determined that the target vehicle's steering intention is to go straight, or, when the roll angle is equal to 100°, it is determined that the target vehicle's steering intention is to turn right.

在一种实现方式中，目标车辆的侧倾运动可以用公式(3)表示。In one implementation, the roll motion of the target vehicle can be expressed by formula (3).

公式(3)中，

为侧倾角，

为对

进行二次求导的运算，σ为当前轨迹曲率，v为目标车辆质心处的纵向速度，g为重力加速度，h为质心高度，b为质心到后轴的距离，ψ为航向角，

为对ψ进行二次求导的运算。将公式(3)按照hσ＜＜1进行简化假设，可以推导得到公式(4)。 In formula (3),

is the roll angle,

for right

Carry out the calculation of the second derivation, σ is the curvature of the current trajectory, v is the longitudinal velocity at the center of mass of the target vehicle, g is the acceleration of gravity, h is the height of the center of mass, b is the distance from the center of mass to the rear axle, ψ is the heading angle,

is the operation of performing the second derivative on ψ. Simplify formula (3) assuming that hσ<<1, formula (4) can be derived.

公式(4)中，

为目标车辆的准静态侧偏角，指目标车辆在稳态转向下的侧倾角，a _l为两轮质心处的横向加速度，g为重力加速度，具体表达为公式(5)。 In formula (4),

is the quasi-static slip angle of the target vehicle, which refers to the roll angle of the target vehicle under steady-state steering, a _l is the lateral acceleration at the center of mass of the two wheels, and g is the gravitational acceleration, specifically expressed as formula (5).

公式(5)中，a _l为两轮质心处的横向加速度，σ为当前轨迹曲率，v为目标车辆质心处的纵向速度，b为质心到后轴的距离，ψ为航向角，

为对ψ进行二次求导的运算。 In formula (5), a _l is the lateral acceleration at the center of mass of the two wheels, σ is the curvature of the current trajectory, v is the longitudinal velocity at the center of mass of the target vehicle, b is the distance from the center of mass to the rear axle, ψ is the heading angle,

is the operation of performing the second derivative on ψ.

公式(4)可以看出，目标车辆的侧倾角与目标车辆的横向加速度有直接联系，目标车辆的横向加速度能够表征目标车辆的转向意图，从而可知，目标车辆的侧倾角能够确定目标车辆的转向意图。It can be seen from formula (4) that the roll angle of the target vehicle is directly related to the lateral acceleration of the target vehicle, and the lateral acceleration of the target vehicle can represent the steering intention of the target vehicle, so it can be seen that the roll angle of the target vehicle can determine the steering direction of the target vehicle intention.

综上所述，本申请实施例可以通过自动驾驶车辆上的摄像头或者集成了摄像功能的智能传感器，从视觉图像中分析估计目标车辆当前的运动姿态，并据此识别目标车辆此刻的转向意图，最终输入给自动驾驶决策模块以做出合理的避障措施。To sum up, the embodiment of the present application can analyze and estimate the current motion posture of the target vehicle from the visual image through the camera on the self-driving vehicle or the smart sensor integrated with the camera function, and accordingly identify the steering intention of the target vehicle at the moment, Finally, it is input to the automatic driving decision-making module to make reasonable obstacle avoidance measures.

在一个可能的实施例中，如图11所示，本申请实施例通过自动驾驶车辆的一个模块实现。该模块的输入为车辆的目标检测和追踪结果，包括目标车辆视觉图像以及目标车辆在三维空间坐标系中的行驶速度和方向。该模块的输出是将识别得到的转向意图输出给自动驾驶***的行为规划模块，以供自动驾驶车辆制定安全合理的驾驶策略。本申请实施例的模块包含车轮检测、姿态估计和转向意图识别三个子模块。In a possible embodiment, as shown in Figure 11, the embodiment of this application is realized by a module of an autonomous vehicle. The input of this module is the target detection and tracking results of the vehicle, including the visual image of the target vehicle and the driving speed and direction of the target vehicle in the three-dimensional space coordinate system. The output of this module is to output the recognized steering intention to the behavior planning module of the automatic driving system, so that the automatic driving vehicle can formulate a safe and reasonable driving strategy. The module of the embodiment of the present application includes three sub-modules of wheel detection, pose estimation and steering intention recognition.

其中，车轮检测子模块，用于获取目标车辆的图像，还用于获取目标车辆的位置信息，例如目标车辆相对自动驾驶车辆的距离。根据获取的目标车辆图像，利用弧支持椭圆检测算法提取所有的椭圆特征，利用多层次约束的方法从所有椭圆特征中优选出车轮椭圆。基于车轮视觉特征分析建立车轮中心位置约束、尺寸约束、极性约束，用以从所有椭圆特征中剔除噪声椭圆并优选出车轮椭圆。车轮检测结果为车轮椭圆的几何参数，输出给姿态估计子模块。Wherein, the wheel detection sub-module is used to obtain the image of the target vehicle, and is also used to obtain the position information of the target vehicle, such as the distance of the target vehicle from the automatic driving vehicle. According to the obtained target vehicle image, all the ellipse features are extracted by using the arc-supported ellipse detection algorithm, and the wheel ellipse is optimized from all the ellipse features by using the multi-level constraint method. Based on the analysis of wheel visual features, the wheel center position constraints, size constraints, and polarity constraints are established to remove noise ellipses from all ellipse features and optimize wheel ellipses. The wheel detection result is the geometric parameter of the wheel ellipse, which is output to the attitude estimation sub-module.

姿态估计子模块，用于将从车轮检测子模块得到的车轮椭圆利用相机投影模型转换为三维相机坐标系中的椭圆锥面，从椭圆锥面中利用空间圆姿态测量方法求解得到车轮在三维相机坐标系中所在平面或者车轮在三维相机坐标系中的法向量。将前后轮法向量夹角作为转向角，后轮与地面法向量的夹角作为侧倾角，前后轮之间的空间距离作为轴距，输出给后续的转向意图识别子模块。示例地，当目标车辆包括前轮和一个后轮时，所述轴距为前后轮之间的空间距离作为轴距，当目标车辆包括前轮和多个后轮时，所述轴距为前轮和最后一个后轮之间的空间距离作为轴距。The attitude estimation sub-module is used to convert the wheel ellipse obtained from the wheel detection sub-module into an ellipse cone in the three-dimensional camera coordinate system using the camera projection model, and obtain the wheel in the three-dimensional camera by using the space circle attitude measurement method from the ellipse cone. The normal vector of the plane in the coordinate system or the wheel in the 3D camera coordinate system. The angle between the front and rear wheel normal vectors is used as the steering angle, the angle between the rear wheels and the ground normal vector is used as the roll angle, and the space distance between the front and rear wheels is used as the wheelbase, and is output to the subsequent steering intention recognition sub-module. For example, when the target vehicle includes front wheels and one rear wheel, the wheelbase is the space distance between the front and rear wheels as the wheelbase; when the target vehicle includes front wheels and a plurality of rear wheels, the wheelbase is the front wheel The space distance between the wheel and the last rear wheel is taken as the wheelbase.

转向意图识别子模块，用于基于车辆运动特征分析，对目标的侧倾角进行解析，推理得到目标当前时刻的转向意图，将其最终输出给行为规划模块以制定安全合理的驾驶策略。本申请基于两轮车辆简化动力学模型经过仿真实验和理论推导，得到车辆侧倾角与横向加速度和转向方向的关系，进而将其应用到转向意图识别当中。The steering intention recognition sub-module is used to analyze the roll angle of the target based on the analysis of the vehicle motion characteristics, infer the current steering intention of the target, and finally output it to the behavior planning module to formulate a safe and reasonable driving strategy. Based on the simplified dynamic model of the two-wheeled vehicle, the application obtains the relationship between the vehicle roll angle, the lateral acceleration and the steering direction through simulation experiments and theoretical derivation, and then applies it to the steering intention recognition.

基于同一技术构思，本申请实施例还提供了一种目标车辆转向意图的确定装置，该装置可以用于实现本申请实施例提供的目标车辆转向意图的确定方法。Based on the same technical idea, the embodiment of the present application also provides a device for determining the steering intention of the target vehicle, which can be used to implement the method for determining the steering intention of the target vehicle provided in the embodiment of the present application.

基于上述内容和相同构思，如图12所示，本申请还提供一种目标车辆转向意图的确定装置，用于实现以上方法实施例部分介绍的目标车辆转向意图的确定方法，具备上述方法实施例所具备的有益效果。Based on the above content and the same idea, as shown in Figure 12, the present application also provides a device for determining the steering intention of the target vehicle, which is used to implement the method for determining the steering intention of the target vehicle introduced in the above method embodiments, with the above method embodiments have beneficial effects.

示例性地，目标车辆转向意图的确定装置1200包括处理模块1201和输入输出模块1202。其中，输入输出模块1202可用于获取目标车辆的视觉信息；处理模块1201，用于确定视觉信息中目标车辆的姿态信息，其中姿态信息为表示目标车辆的车身姿态的信息；以及用于根据目标车辆的姿态信息，确定目标车辆的转向意图。Exemplarily, the device 1200 for determining the steering intention of the target vehicle includes a processing module 1201 and an input-output module 1202 . Among them, the input and output module 1202 can be used to obtain the visual information of the target vehicle; the processing module 1201 is used to determine the posture information of the target vehicle in the visual information, wherein the posture information is information representing the body posture of the target vehicle; attitude information to determine the steering intention of the target vehicle.

输入输出模块1202例如可以是摄像头、摄像***或相机传感器。The input and output module 1202 may be, for example, a camera, a camera system or a camera sensor.

处理模块1201和输入输出模块1202还可以用于执行上述方法实施例对应的其它步骤，在此不再赘述。The processing module 1201 and the input/output module 1202 may also be used to perform other steps corresponding to the above method embodiments, which will not be repeated here.

应理解的是，本申请实施例中的目标车辆转向意图的确定装置可以由软件实现，例如，具有上述功能的计算机程序或指令来实现，相应计算机程序或指令可以存储在目标车辆转向意图的确定装置内部的存储器中，通过处理器读取该存储器内部的相应计算机程序或指令来实现处理模块1201和/或输入输出模块1202的上述功能。或者，本申请实施例中的目标车辆转向意图的确定装置还可以由硬件来实现。其中，处理模块1201可以是处理器(如CPU或***芯片中的处理器)，或者，本申请实施例中的目标车辆转向意图的确定装置还可以由处理器和软件模块的结合实现。It should be understood that the device for determining the steering intention of the target vehicle in the embodiments of the present application may be realized by software, for example, a computer program or instruction having the above functions, and the corresponding computer program or instruction may be stored in the Determination of the steering intention of the target vehicle In the internal memory of the device, the above-mentioned functions of the processing module 1201 and/or the input and output module 1202 are realized by the processor reading the corresponding computer program or instructions in the memory. Alternatively, the device for determining the steering intention of the target vehicle in the embodiment of the present application may also be implemented by hardware. Wherein, the processing module 1201 may be a processor (such as a processor in a CPU or a system chip), or, the device for determining the steering intention of the target vehicle in the embodiment of the present application may also be implemented by a combination of a processor and a software module.

基于同一技术构思，如图13所示，本申请还提供一种目标车辆转向意图的确定装置，用于实现以上方法实施例部分介绍的目标车辆转向意图的确定方法，具备上述方法实施例所具备的有益效果。示例性地，目标车辆转向意图的确定装置1300包括处理器1301、接口电路1302，还可以包括存储器1303。Based on the same technical concept, as shown in Figure 13, the present application also provides a device for determining the steering intention of the target vehicle, which is used to implement the method for determining the steering intention of the target vehicle introduced in the above method embodiments, and has all the features of the above method embodiments beneficial effect. Exemplarily, the device 1300 for determining the steering intention of the target vehicle includes a processor 1301 , an interface circuit 1302 , and may also include a memory 1303 .

其中，接口电路1302可用于获取目标车辆的视觉信息；处理器1301，用于检测视觉信息中目标车辆的姿态信息，其中姿态信息为表示目标车辆的车身姿态的信息；以及用于根据目标车辆的姿态信息，确定目标车辆的转向意图。Among them, the interface circuit 1302 can be used to obtain the visual information of the target vehicle; the processor 1301 is used to detect the posture information of the target vehicle in the visual information, wherein the posture information is information representing the body posture of the target vehicle; attitude information to determine the steering intention of the target vehicle.

接口电路1302例如可以是摄像头、摄像***或相机传感器。The interface circuit 1302 can be, for example, a camera, a camera system, or a camera sensor.

存储器1303可以用于存储处理器1301执行的代码、指令或程序。The memory 1303 can be used to store codes, instructions or programs executed by the processor 1301.

目标车辆转向意图的确定装置1300可以是车辆，也可以是车辆内部的芯片。应理解，虽然图3中仅示出了一个处理器、一个存储器和一个接口电路。目标车辆转向意图的确定装置1300还可以包括更多的处理器、接口电路和存储器。The device 1300 for determining the steering intention of the target vehicle may be a vehicle, or a chip inside the vehicle. It should be understood that although only one processor, one memory and one interface circuit are shown in FIG. 3 . The device 1300 for determining the turning intention of the target vehicle may further include more processors, interface circuits and memories.

其中，接口电路还可以用于目标车辆转向意图的确定装置1300与终端或车辆的其他组件连通，例如存储器或其他处理器等。处理器1301可用于通过接口电路与其他组件进行信号交互。接口电路1302可以是处理器的输入/输出接口。Wherein, the interface circuit can also be used for determining the turning intention of the target vehicle. The device 1300 communicates with the terminal or other components of the vehicle, such as memory or other processors. The processor 1301 can be used to perform signal interaction with other components through an interface circuit. Interface circuit 1302 may be an input/output interface of a processor.

例如，处理器可通过接口电路读取与之耦合的存储器中的计算机程序或指令，并译码和执行这些计算机程序或指令。应理解，这些计算机程序或指令可包括上述功能程序，也可以包括上述车辆控制装置的功能程序。当相应功能程序被处理器译码并执行时，可以使得车辆控制装置实现本申请实施例所提供的车辆控制方法中的方案。For example, the processor can read computer programs or instructions in the memory coupled to it through the interface circuit, and decode and execute these computer programs or instructions. It should be understood that these computer programs or instructions may include the above-mentioned function programs, and may also include the above-mentioned function programs of the vehicle control device. When the corresponding function program is decoded and executed by the processor, the vehicle control device can realize the solution in the vehicle control method provided by the embodiment of the present application.

可选的，这些功能程序存储在目标车辆转向意图的确定装置外部的存储器中，此时目标车辆转向意图的确定装置可以不包括存储器。当上述功能程序被处理器译码并执行时，存储器中临时存放上述功能程序的部分或全部内容。Optionally, these functional programs are stored in a memory outside the device for determining the turning intention of the target vehicle, and at this time, the device for determining the turning intention of the target vehicle may not include a memory. When the above functional program is decoded and executed by the processor, part or all of the content of the above functional program is temporarily stored in the memory.

可选的，这些功能程序存储在目标车辆转向意图的确定装置内部的存储器中。当目标车辆转向意图的确定装置内部的存储器中存储有上述功能程序时，目标车辆转向意图的确定装置可被设置在本申请实施例的目标车辆转向意图的确定装置中。Optionally, these function programs are stored in a memory inside the device for determining the steering intention of the target vehicle. When the above-mentioned function program is stored in the internal memory of the device for determining the steering intention of the target vehicle, the device for determining the steering intention of the target vehicle may be provided in the device for determining the steering intention of the target vehicle in the embodiment of the present application.

可选的，这些功能程序存储在目标车辆转向意图的确定装置外部的存储器中，这些功能程序的其他部分存储在目标车辆转向意图的确定装置内部的存储器中。Optionally, these function programs are stored in a memory outside the device for determining the turning intention of the target vehicle, and other parts of these function programs are stored in a memory inside the device for determining the turning intention of the target vehicle.

应理解，上述处理器可以是一个芯片。例如，该处理器可以是现场可编程门阵列(field programmable gate array，FPGA)，可以是专用集成芯片(application specific integrated circuit，ASIC)，还可以是***芯片(system on chip，SoC)，还可以是中央处理器(central processor unit，CPU)，还可以是网络处理器(network processor，NP)，还可以是数字信号处理电路(digital signal processor，DSP)，还可以是微控制器(micro controller unit，MCU)，还可以是可编程控制器(programmable logic device，PLD)或其他集成芯片。It should be understood that the above processor may be a chip. For example, the processor may be a field programmable gate array (field programmable gate array, FPGA), an application specific integrated circuit (ASIC), or a system chip (system on chip, SoC). It can be a central processor unit (CPU), a network processor (network processor, NP), a digital signal processing circuit (digital signal processor, DSP), or a microcontroller (micro controller unit) , MCU), can also be a programmable controller (programmable logic device, PLD) or other integrated chips.

应注意，本申请实施例中的处理器可以是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法实施例的各步骤可以通过处理器中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器可以是通用处理器、数字信号处理器(DSP)、专用集成电路(ASIC)、现场可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本申请实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本申请实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器，处理器读取存储器中的信息，结合其硬件完成上述方法的步骤。It should be noted that the processor in the embodiment of the present application may be an integrated circuit chip, which has a signal processing capability. In the implementation process, each step of the above-mentioned method embodiments may be completed by an integrated logic circuit of hardware in a processor or instructions in the form of software. The above-mentioned processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components . Various methods, steps, and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like. The steps of the methods disclosed in connection with the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, register. The storage medium is located in the memory, and the processor reads the information in the memory, and completes the steps of the above method in combination with its hardware.

可以理解，本申请实施例中的存储器可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(read-only memory，ROM)、可编程只读存储器(programmable ROM，PROM)、可擦除可编程只读存储器(erasable PROM，EPROM)、电可擦除可编程只读存储器(electrically EPROM，EEPROM)或闪存。易失性存储器可以是随机存取存储器(random access memory，RAM)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(static RAM，SRAM)、动态随机存取存储器(dynamic RAM，DRAM)、同步动态随机存取存储器(synchronous DRAM，SDRAM)、双倍数据速率同步动态随机存取存储器(double data rate SDRAM，DDR SDRAM)、增强型同步动态随机存取存储器(enhanced SDRAM，ESDRAM)、同步连接动态随机存取存储器(synchlink DRAM，SLDRAM)和直接内存总线随机存取存储器(direct rambus RAM，DR RAM)。应注意，本文描述的***和方法的存储器旨在包括但不限于这些和任意其它适合类型的存储器。It can be understood that the memory in the embodiments of the present application may be a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memories. Among them, the non-volatile memory can be read-only memory (read-only memory, ROM), programmable read-only memory (programmable ROM, PROM), erasable programmable read-only memory (erasable PROM, EPROM), electrically programmable Erases programmable read-only memory (electrically EPROM, EEPROM) or flash memory. Volatile memory can be random access memory (RAM), which acts as external cache memory. By way of illustration and not limitation, many forms of RAM are available such as static random access memory (static RAM, SRAM), dynamic random access memory (dynamic RAM, DRAM), synchronous dynamic random access memory (synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (double data rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous connection dynamic random access memory (synchlink DRAM, SLDRAM ) and direct memory bus random access memory (direct rambus RAM, DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but not be limited to, these and any other suitable types of memory.

应理解，在通过图13所示结构实现该目标车辆转向意图的确定装置时，可由存储器存储计算机程序或指令，由处理器执行存储器中存储的计算机程序或指令，执行在通过图12所示结构实现该目标车辆转向意图的确定装置时由处理模块1201执行的动作。还可由接口电路执行在通过图12所示结构实现该目标车辆转向意图的确定装置时由输入输出模块1202执行的动作。可选的，可由图13所示的处理器和存储器实现图12所示的处理模块1201，或者说，图12所示的处理模块1201包括处理器和存储器，或者说，由处理器执行存储器中存储的计算机程序或指令，实现由以上图12所示处理模块1201执行的动作。和/或，可由图13所示的接口电路实现图12所示的输入输出模块1202，或者说，图12所示的处理模块1201包括图13所示的接口电路，或者说，由接口电路执行以上图12所示输入输出模块1202执行的动作。It should be understood that when the device for determining the steering intention of the target vehicle is realized through the structure shown in FIG. Actions performed by the processing module 1201 when the device for determining the steering intention of the target vehicle is implemented. The actions performed by the input-output module 1202 when the device for determining the steering intention of the target vehicle is implemented through the structure shown in FIG. 12 can also be performed by the interface circuit. Optionally, the processing module 1201 shown in FIG. 12 may be implemented by the processor and memory shown in FIG. 13, or in other words, the processing module 1201 shown in FIG. The stored computer programs or instructions implement the actions performed by the processing module 1201 shown in FIG. 12 above. And/or, the input and output module 1202 shown in FIG. 12 can be realized by the interface circuit shown in FIG. 13, or in other words, the processing module 1201 shown in FIG. 12 includes the interface circuit shown in FIG. The actions performed by the input and output module 1202 are shown in FIG. 12 above.

应理解，图12和图13任一所示的目标车辆转向意图的确定装置的结构可以互相结合，图12和图13任一所示的目标车辆转向意图的确定装置以及各可选实施例相关设计细节可互相参考，也可以参考图12和图13任一所示的目标车辆转向意图的确定方法以及各可选实施例相关设计细节。此处不再重复赘述。It should be understood that the structure of the device for determining the steering intention of the target vehicle shown in any one of Fig. 12 and Fig. 13 can be combined with each other, and the device for determining the steering intention of the target vehicle shown in any one of Fig. 12 and Fig. The design details can refer to each other, and can also refer to the method for determining the steering intention of the target vehicle shown in any one of FIG. 12 and FIG. 13 and the relevant design details of each alternative embodiment. It will not be repeated here.

基于上述内容和相同技术构思，本申请提供一种计算设备，包括处理器，处理器与存储器相连，存储器用于存储计算机程序或指令，处理器用于执行存储器中存储的计算机程序，以使得计算设备执行上述方法实施例中的方法。Based on the above content and the same technical concept, the present application provides a computing device, including a processor, the processor is connected to a memory, the memory is used to store computer programs or instructions, and the processor is used to execute the computer program stored in the memory, so that the computing device Execute the methods in the above method embodiments.

基于上述内容和相同技术构思，本申请提供一种计算机可读存储介质，其上存储有计算机程序或指令，当该计算机程序或指令被执行时，以使得计算设备执行上述方法实施例中的方法。Based on the above content and the same technical concept, the present application provides a computer-readable storage medium, on which a computer program or instruction is stored, and when the computer program or instruction is executed, the computing device executes the method in the above method embodiment .

基于上述内容和相同技术构思，本申请提供一种计算机程序产品，当计算机执行计算机程序产品时，以使得计算设备执行上述方法实施例中的方法。Based on the above content and the same technical idea, the present application provides a computer program product, which enables the computing device to execute the methods in the above method embodiments when the computer executes the computer program product.

基于上述内容和相同技术构思，本申请提供一种芯片，芯片与存储器相连，用于读取并执行存储器中存储的计算机程序或指令，以使得计算设备执行上述方法实施例中的方法。Based on the above content and the same technical concept, the present application provides a chip, which is connected to a memory, and is used to read and execute computer programs or instructions stored in the memory, so that the computing device executes the methods in the above method embodiments.

基于上述内容和相同构思，本申请实施例提供一种装置，所述装置包括处理器和接口电路，所述接口电路，用于接收计算机程序或指令并传输至所述处理器；所述处理器运行所述计算机程序或指令以执行上述方法实施例中的方法。Based on the above content and the same idea, an embodiment of the present application provides a device, the device includes a processor and an interface circuit, the interface circuit is used to receive computer programs or instructions and transmit them to the processor; the processor Execute the computer program or instructions to execute the methods in the above method embodiments.

应理解，本申请实施例中对模块的划分是示意性的，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式。另外，在本申请各个实施例中的各功能模块可以集成在一个处理器中，也可以是单独物理存在，也可以两个或两个以上模块集成在一个模块中。上述集成的模块既可以采用硬件的形式实现，也可以采用软件功能模块的形式实现。It should be understood that the division of modules in the embodiment of the present application is schematic, and is only a logical function division, and there may be another division manner in actual implementation. In addition, each functional module in each embodiment of the present application may be integrated into one processor, or physically exist separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software function modules.

本领域内的技术人员应明白，本申请的实施例可提供为方法、***、或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

本申请是参照根据本申请实施例的方法、设备(***)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。The present application is described with reference to flowcharts and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow diagram procedure or procedures and/or block diagram procedures or blocks.

尽管已描述了本申请的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例作出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本申请范围的所有变更和修改。While preferred embodiments of the present application have been described, additional changes and modifications to these embodiments can be made by those skilled in the art once the basic inventive concept is appreciated. Therefore, the appended claims are intended to be construed to cover the preferred embodiment and all changes and modifications which fall within the scope of the application.

显然，本领域的技术人员可以对本申请实施例进行各种改动和变型而不脱离本申请实施例的精神和范围。这样，倘若本申请实施例的这些修改和变型属于本申请权利要求及其等同技术的范围之内，则本申请也意图包含这些改动和变型在内。Apparently, those skilled in the art can make various changes and modifications to the embodiments of the present application without departing from the spirit and scope of the embodiments of the present application. In this way, if the modifications and variations of the embodiments of the present application fall within the scope of the claims of the present application and equivalent technologies, the present application also intends to include these modifications and variations.

Claims

一种目标车辆转向意图的确定方法，其特征在于，包括：A method for determining the steering intention of a target vehicle, comprising:

获取所述目标车辆的视觉信息；acquiring visual information of the target vehicle;

根据所述视觉信息确定所述目标车辆的姿态信息，所述姿态信息为表示所述目标车辆的车身姿态的信息；determining attitude information of the target vehicle according to the visual information, where the attitude information is information representing the body attitude of the target vehicle;

根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图。According to the posture information of the target vehicle, the steering intention of the target vehicle is determined.
如权利要求1所述的方法，其特征在于，所述姿态信息包括所述目标车辆的侧倾角。The method of claim 1, wherein the attitude information includes a roll angle of the target vehicle.
如权利要求2所述的方法，其特征在于，根据所述目标车辆的所述侧倾角，确定所述目标车辆的转向意图，包括：The method according to claim 2, wherein determining the steering intention of the target vehicle according to the roll angle of the target vehicle comprises:

当所述侧倾角小于第一阈值时，确定所述目标车辆的转向意图为左转；When the roll angle is less than a first threshold, it is determined that the steering intention of the target vehicle is to turn left;

当所述侧倾角大于所述第一阈值小于第二阈值时，确定所述目标车辆的转向意图为直行；或When the roll angle is greater than the first threshold and less than a second threshold, determining that the steering intention of the target vehicle is to go straight; or

当所述侧倾角大于所述第二阈值时，确定所述目标车辆的转向意图为右转；When the roll angle is greater than the second threshold, it is determined that the steering intention of the target vehicle is a right turn;

其中，所述第一阈值为负值，所述第二阈值为正值。Wherein, the first threshold is a negative value, and the second threshold is a positive value.
如权利要求2或3所述的方法，其特征在于，The method according to claim 2 or 3, characterized in that,

所述目标车辆的所述侧倾角通过以下参数确定：所述目标车辆的空间法向量、以及地面法向量。The roll angle of the target vehicle is determined by the following parameters: a space normal vector of the target vehicle and a ground normal vector.
如权利要求4所述的方法，其特征在于，所述侧倾角符合以下公式：The method according to claim 4, wherein the roll angle complies with the following formula:

其中，
为所述侧倾角，n _r为所述目标车辆的空间法向量，n _g为所述地面法向量。
in,
is the roll angle, n _r is the space normal vector of the target vehicle, and n _g is the ground normal vector.
如权利要求1至5任一项所述的方法，其特征在于，The method according to any one of claims 1 to 5, characterized in that,

所述目标车辆的姿态信息还包括所述目标车辆的转向角，所述转向角为所述目标车辆的前车轮的偏转角度。The attitude information of the target vehicle also includes the steering angle of the target vehicle, and the steering angle is the deflection angle of the front wheels of the target vehicle.
如权利要求6所述的方法，其特征在于，所述目标车辆的所述转向角符合下述公式：The method according to claim 6, wherein the steering angle of the target vehicle complies with the following formula:

其中，δ为所述转向角，n _f为所述目标车辆的所述前车轮的空间法向量；n _r为所述目标车辆的空间法向量。
Wherein, δ is the steering angle, n _f is the space normal vector of the front wheel of the target vehicle; n _r is the space normal vector of the target vehicle.
如权利要求1至7任一项所述的方法，其特征在于，所述根据所述视觉信息确定所述目标车辆的姿态信息，包括：The method according to any one of claims 1 to 7, wherein the determining the attitude information of the target vehicle according to the visual information comprises:

确定所述目标车辆在所述视觉信息中对应的几何形状；determining a geometric shape corresponding to the target vehicle in the visual information;

根据所述几何形状确定所述目标车辆的姿态信息。The attitude information of the target vehicle is determined according to the geometric shape.
如权利要求8所述的方法，其特征在于，所述确定所述目标车辆在所述视觉信息中对应的几何形状，包括：The method according to claim 8, wherein the determining the geometric shape corresponding to the target vehicle in the visual information comprises:

确定所述目标车辆的车轮在所述视觉信息中对应的第一几何形状。A first geometric shape corresponding to a wheel of the target vehicle in the visual information is determined.
如权利要求9所述的方法，其特征在于，当所述目标车辆的车轮在所述视觉信息中的所述第一几何形状为椭圆时，第一椭圆为在所述视觉信息中的多个椭圆中、尺寸占比满足设定的比例范围的椭圆，所述第一椭圆为所述目标车辆的车轮在所述视觉信息中的椭圆。The method according to claim 9, wherein when the first geometric shape of the wheel of the target vehicle in the visual information is an ellipse, the first ellipse is a plurality of geometric shapes in the visual information. Among the ellipses, the ellipse whose size ratio satisfies the set ratio range, the first ellipse is the ellipse of the wheel of the target vehicle in the visual information.
如权利要求10所述的方法，其特征在于，所述约定条件还包括：The method according to claim 10, wherein the agreed conditions further comprise:

所述第一椭圆的中心位置在所述视觉信息的水平中心线的下方，所述水平中心线将所述视觉信息分为上下两部分；和/或，The central position of the first ellipse is below the horizontal centerline of the visual information, and the horizontal centerline divides the visual information into upper and lower parts; and/or,

所述第一椭圆的边缘的颜色符合：边缘外侧的颜色深于边缘内侧的颜色。The color of the edge of the first ellipse is as follows: the color outside the edge is darker than the color inside the edge.
如权利要求1至8任一项所述的方法，其特征在于，所述目标车辆的姿态信息还包括所述目标车辆的车身长度，所述方法还包括：The method according to any one of claims 1 to 8, wherein the attitude information of the target vehicle also includes the body length of the target vehicle, and the method further comprises:

根据所述目标车辆的所述车身长度，确定所述目标车辆的转向意图。According to the body length of the target vehicle, the steering intention of the target vehicle is determined.
如权利要求1至12任一所述的方法，其特征在于，The method according to any one of claims 1 to 12, characterized in that,

所述目标车辆为直排轮车。The target vehicle is a straight-wheeled vehicle.
如权利要求1至13所述的方法，其特征在于，The method according to claims 1 to 13, characterized in that,

所述目标车辆为自行车，或者电动车，或者燃油车。The target vehicle is a bicycle, or an electric vehicle, or a fuel vehicle.
一种安全驾驶方法，其特征在于，所述安全驾驶方法包括：A safe driving method, characterized in that the safe driving method comprises:

根据权利要求1至14任一所述转向意图的确定方法，确定安全驾驶策略。According to the method for determining a steering intention according to any one of claims 1 to 14, a safe driving strategy is determined.
一种目标车辆转向意图的确定装置，其特征在于，包括：A device for determining the steering intention of a target vehicle, characterized in that it includes:

输入输出模块，用于获取目标车辆的视觉信息；The input and output module is used to obtain the visual information of the target vehicle;

处理模块，用于确定所述视觉信息中所述目标车辆的姿态信息，所述姿态信息为表示所述目标车辆的车身姿态的信息；以及用于根据所述目标车辆的姿态信息，确定所述目标车辆的转向意图。A processing module, configured to determine attitude information of the target vehicle in the visual information, the attitude information being information representing the body attitude of the target vehicle; and determining the attitude information of the target vehicle according to the attitude information of the target vehicle. The steering intent of the target vehicle.
如权利要求16所述的装置，其特征在于，所述姿态信息包括所述目标车辆的侧倾角。The apparatus of claim 16, wherein the attitude information includes a roll angle of the target vehicle.
如权利要求17所述的装置，其特征在于，在根据所述目标车辆的所述侧倾角，确定所述目标车辆的转向意图时，所述处理模块具体用于：The device according to claim 17, wherein when determining the steering intention of the target vehicle according to the roll angle of the target vehicle, the processing module is specifically configured to:

当所述侧倾角小于第一阈值时，确定所述目标车辆的转向意图为左转；When the roll angle is less than a first threshold, it is determined that the steering intention of the target vehicle is to turn left;

当所述侧倾角大于所述第一阈值小于第二阈值时，确定所述目标车辆的转向意图为直行；或When the roll angle is greater than the first threshold and less than a second threshold, determining that the steering intention of the target vehicle is to go straight; or

当所述侧倾角大于所述第二阈值时，确定所述目标车辆的转向意图为右转；When the roll angle is greater than the second threshold, it is determined that the steering intention of the target vehicle is a right turn;

其中，所述第一阈值为负值，所述第二阈值为正值。Wherein, the first threshold is a negative value, and the second threshold is a positive value.
如权利要求17或18所述的装置，其特征在于，Apparatus as claimed in claim 17 or 18, characterized in that,

所述目标车辆的所述侧倾角通过以下参数确定：所述目标车辆的空间法向量、以及地面法向量。The roll angle of the target vehicle is determined by the following parameters: a space normal vector of the target vehicle and a ground normal vector.
如权利要求19所述的装置，其特征在于，所述侧倾角符合以下公式：The device according to claim 19, wherein the roll angle conforms to the following formula:

其中，
为所述侧倾角，n _r为所述目标车辆的空间法向量，n _g为所述地面法向量。
in,
is the roll angle, n _r is the space normal vector of the target vehicle, and n _g is the ground normal vector.
如权利要求17至20任一项所述的装置，其特征在于，Apparatus according to any one of claims 17 to 20, characterized in that

所述目标车辆的姿态信息还包括所述目标车辆的转向角，所述转向角为所述目标车辆的前车轮的偏转角度。The attitude information of the target vehicle also includes the steering angle of the target vehicle, and the steering angle is the deflection angle of the front wheels of the target vehicle.
如权利要求21所述的装置，其特征在于，所述目标车辆的所述转向角符合下述公式：The device according to claim 21, wherein the steering angle of the target vehicle complies with the following formula:

其中，δ为所述转向角，n _f为所述目标车辆的所述前车轮的空间法向量；n _r为所述目标车辆的空间法向量。
Wherein, δ is the steering angle, n _f is the space normal vector of the front wheel of the target vehicle; n _r is the space normal vector of the target vehicle.
如权利要求17至22任一项所述的装置，其特征在于，在根据所述视觉信息确定所述目标车辆的姿态信息时，所述处理模块具体用于：The device according to any one of claims 17 to 22, wherein when determining the attitude information of the target vehicle according to the visual information, the processing module is specifically configured to:

确定所述目标车辆在所述视觉信息中对应的几何形状；determining a geometric shape corresponding to the target vehicle in the visual information;

根据所述几何形状确定所述目标车辆的姿态信息。The attitude information of the target vehicle is determined according to the geometric shape.
如权利要求23所述的装置，其特征在于，在确定所述目标车辆在所述视觉信息中对应的几何形状时，所述处理模块具体用于：The device according to claim 23, wherein when determining the geometric shape corresponding to the target vehicle in the visual information, the processing module is specifically configured to:

确定所述目标车辆的车轮在所述视觉信息中对应的第一几何形状。A first geometric shape corresponding to a wheel of the target vehicle in the visual information is determined.
如权利要求24所述的装置，其特征在于，当所述目标车辆的车轮在所述视觉信息中的所述第一几何形状为椭圆时，The device according to claim 24, wherein when the first geometric shape of the wheel of the target vehicle in the visual information is an ellipse,

第一椭圆为在所述视觉信息中的多个椭圆中、尺寸占比满足设定的比例范围的椭圆，其中，所述第一椭圆为所述目标车辆的车轮在所述视觉信息中的椭圆。The first ellipse is an ellipse whose size ratio satisfies a set ratio range among multiple ellipses in the visual information, wherein the first ellipse is an ellipse of a wheel of the target vehicle in the visual information .
如权利要求25所述的装置，其特征在于，所述约定条件还包括：The device according to claim 25, wherein the agreed conditions further include:

所述第一椭圆的中心位置在所述视觉信息的水平中心线的下方，所述水平中心线将所述视觉信息分为上下两部分；和/或，The center position of the first ellipse is below the horizontal centerline of the visual information, and the horizontal centerline divides the visual information into upper and lower parts; and/or,

所述第一椭圆的边缘的颜色符合：边缘外侧的颜色深于边缘内侧的颜色。The color of the edge of the first ellipse is as follows: the color outside the edge is darker than the color inside the edge.
如权利要求17至22任一项所述的装置，其特征在于，所述目标车辆的姿态信息还包括所述目标车辆的车身长度，所述处理模块还用于：The device according to any one of claims 17 to 22, wherein the attitude information of the target vehicle also includes the body length of the target vehicle, and the processing module is further used for:

根据所述目标车辆的所述车身长度，确定所述目标车辆的转向意图。According to the body length of the target vehicle, the steering intention of the target vehicle is determined.
如权利要求17至27任一所述的装置，其特征在于，Apparatus according to any one of claims 17 to 27, wherein

所述目标车辆为直排轮车。The target vehicle is a straight-wheeled vehicle.
如权利要求17至28所述的装置，其特征在于，Apparatus according to claims 17 to 28, characterized in that

所述目标车辆为自行车，或者电动车，或者燃油车。The target vehicle is a bicycle, or an electric vehicle, or a fuel vehicle.
一种安全驾驶装置，其特征在于，包括：A safety driving device, characterized in that it comprises:

输入输出模块，用于获取目标车辆的视觉信息；The input and output module is used to obtain the visual information of the target vehicle;

处理模块，用于根据如权利要求1至14任一所述的转向意图的确定方法，确定安全驾驶策略。A processing module, configured to determine a safe driving strategy according to the method for determining a steering intention according to any one of claims 1 to 14.
一种计算设备，其特征在于，包括处理器，所述处理器与存储器相连，所述存储器存储计算机程序或指令，所述处理器用于执行所述存储器中存储的计算机程序或指令，以使得所述计算设备执行如权利要求1至15中任一项所述的方法。A computing device, characterized in that it includes a processor, the processor is connected to a memory, the memory stores computer programs or instructions, and the processor is used to execute the computer programs or instructions stored in the memory, so that the The computing device performs the method according to any one of claims 1-15.
一种计算机可读存储介质，其特征在于，所述计算机可读存储介质中存储有计算机程序或指令，当所述计算机程序或指令被计算设备执行时，以使得所述计算设备执行如权利要求1至15中任一项所述的方法。A computer-readable storage medium, characterized in that, computer programs or instructions are stored in the computer-readable storage medium, and when the computer programs or instructions are executed by a computing device, the computing device performs the The method described in any one of 1 to 15.
一种芯片，其特征在于，包括至少一个处理器和接口；A chip, characterized in that it includes at least one processor and an interface;

所述接口，用于为所述至少一个处理器提供计算机程序、指令或者数据；said interface for providing said at least one processor with computer programs, instructions or data;

所述至少一个处理器用于执行所述计算机程序或指令，以使得如权利要求1至15中任一项所述的方法被执行。The at least one processor is configured to execute the computer program or instructions such that the method as claimed in any one of claims 1 to 15 is performed.