WO2020019930A1

WO2020019930A1 - Automatic parking method and device

Info

Publication number: WO2020019930A1
Application number: PCT/CN2019/093421
Authority: WO
Inventors: 张雪飞; 李良; 蒋少峰
Original assignee: 广州小鹏汽车科技有限公司
Priority date: 2018-07-25
Filing date: 2019-06-28
Publication date: 2020-01-30
Also published as: CN110758246A; CN110758246B

Abstract

An automatic parking method and device, comprising: receiving a parking instruction; performing perspective transformation on peripheral images collected by an acquisition camera; inputting the transformed images into an idle parking space feature model; identifying areas in the images that comply with an idle parking space feature, and determining an inclination angle and location information in the images of each area; further determining the location of each idle parking space relative to a vehicle; receiving a selection instruction for a target parking space; planning a parking route for the vehicle according to the inclination angle and the location of the target parking space relative to the vehicle; and parking according to the parking route.

Description

一种自动泊车方法及装置 Automatic parking method and device Ranch

技术领域Technical field

本申请涉及车辆技术领域，尤其涉及一种自动泊车方法及装置。The present application relates to the technical field of vehicles, and in particular, to an automatic parking method and device.

背景技术Background technique

随着人工智能的不断发展，车辆的智能程度也不断提升，自动泊车因为具有广泛的应用前景而备受瞩目。With the continuous development of artificial intelligence, the intelligence of vehicles has also been continuously improved, and automatic parking has attracted much attention because of its wide application prospects.

在进行自动泊车时，寻找空闲车位是非常重要的环节，目前的自动泊车***均是依靠车辆上部署的雷达***来寻找空闲车位，而雷达***是依靠探测障碍物来发现空闲车位，比如，当两个被占车位中间有一个空闲车位时，雷达***可探测到被占车位上车辆的位置（即空闲车位左右两侧车辆的位置），之后，可分析这两个车辆的位置来确定空闲车位的位置，而对于图1所示的连续的、或成片的空闲车位，因为空闲车位附近没有已停车辆作为参考，雷达***则无能为力。In automatic parking, finding free parking spaces is a very important part. At present, automatic parking systems rely on radar systems deployed on vehicles to find free parking spaces, and radar systems rely on detecting obstacles to find free parking spaces, such as When there is a free parking space between the two occupied parking spaces, the radar system can detect the position of the vehicle on the occupied parking space (ie, the positions of the left and right sides of the free parking space), and then the positions of the two vehicles can be analyzed to determine The location of the free parking space. For the continuous or piece of free parking space shown in Figure 1, the radar system is powerless because there are no parked vehicles near the free parking space as a reference.

发明内容Summary of the Invention

本申请实施例提供一种自动泊车方法及装置，用以解决现有技术中雷达***不能探测到附近没有障碍物的空闲车位而无法进行自动泊车的问题。The embodiments of the present application provide an automatic parking method and device, which are used to solve the problem that the radar system cannot detect an idle parking space without obstacles in the prior art and cannot perform automatic parking.

第一方面，本申请实施例提供的一种自动泊车方法，包括：In a first aspect, an automatic parking method provided by an embodiment of the present application includes:

接收泊车指令，获取摄像头采集到的周围图像，对所述图像进行透视变换；Receiving a parking instruction, obtaining a surrounding image collected by a camera, and performing perspective transformation on the image;

将进行透视变换后的图像输入到预先训练的空闲车位特征模型中，利用所述空闲车位特征模型识别出所述图像中符合空闲车位特征的区域，并确定每一区域的倾角和在所述图像中的位置信息，其中，每一区域代表一个空闲车位，区域的倾角为空闲车位的倾角，空闲车位的倾角是指选定的车位中心线与基准线之间的夹角；The perspective-transformed image is input into a pre-trained free parking space feature model, and the free parking space feature model is used to identify areas in the image that conform to the free parking space feature, and determine the inclination of each area and the image in the image. Position information in the system, where each area represents an idle parking space, the inclination of the area is the inclination of the idle parking space, and the inclination of the idle parking space refers to the angle between the selected parking space centerline and the reference line;

针对每一空闲车位，根据预先标定的所述摄像头采集到的图像中每一像素点相对于所述车辆的位置关系和代表所述空闲车位的区域在所述图像中的位置，确定所述空闲车位相对于所述车辆的位置； For each vacant parking space, determine the vacancy according to the positional relationship of each pixel in the image collected by the camera pre-calibrated relative to the vehicle and the position of the area representing the vacant parking space in the image. The position of the parking space relative to the vehicle; Ranch

接收目标车位的选择指令，根据所述目标车位的倾角和相对于所述车辆的位置为所述车辆规划泊车线路，并根据规划的泊车线路驱动所述车辆进行泊车。Receive a selection instruction of a target parking space, plan a parking line for the vehicle according to the inclination of the target parking space and a position relative to the vehicle, and drive the vehicle to park according to the planned parking line.

本申请实施例中，预先训练空闲车位特征模型，之后，在进行自动泊车时，获取摄像头采集到的周围图像，对周围图像进行透视变换，将进行透视变换后的图输入到空闲车位特征模型中，识别出图像中符合空闲车位特征的区域，并确定每一区域的倾角和在图像中的位置信息，其中，每一区域代表一个空闲车位，这种属于纯视觉的空闲车位识别方式脱离了对障碍物的依赖，因此，对于包含空闲车位的图像，即便空闲车位的附近没有障碍物也可较好地识别出图像中包含的空闲车位的位置和倾角，进而完成自动泊车。In the embodiment of the present application, the free parking space feature model is pre-trained, and then, when the automatic parking is performed, the surrounding images collected by the camera are obtained, the surrounding images are perspectively transformed, and the perspective transformed map is input to the free parking space feature model. In the image, identify the areas in the image that meet the characteristics of free parking spaces, and determine the inclination of each area and the position information in the image. Among them, each area represents a free parking space, which is a purely visual method of identifying free parking spaces. Dependence on obstacles. Therefore, for images containing free parking spaces, even if there are no obstacles near the free parking spaces, the position and inclination of the free parking spaces included in the image can be better identified, and automatic parking can be completed.

在具体实施时，为了提高对空闲车位的识别准确度，还可对空闲车位特征模型识别出的区域进行筛选，因此，在一种可能的实施方式下，在确定每一空闲车位相对于车辆的位置之前之后，还包括：In specific implementation, in order to improve the recognition accuracy of free parking spaces, the areas identified by the free parking space feature model can also be filtered. Therefore, in a possible implementation manner, in determining the Before and after the location, it also includes:

确定所述空闲车位符合预先确定的车位形状规则。It is determined that the free parking space conforms to a predetermined parking space shape rule.

在一种可能的实施方式下，确定所述空闲车位符合预先确定的车位形状规则，包括：In a possible implementation manner, determining that the idle parking space conforms to a predetermined parking space shape rule includes:

选取所述图像中代表所述空闲车位两个相邻顶点的像素点；Selecting pixels in the image that represent two adjacent vertices of the free parking space;

根据预先标定的每一像素点相对于所述车辆的位置，确定所述空闲车位两个相邻顶点分别相对于所述车辆的位置；Determining the positions of two adjacent vertices of the free parking space relative to the vehicle according to the positions of each pixel relative to the vehicle in advance;

根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置，确定所述空闲车位符合预先确定的车位形状规则；或者，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置和所述空闲车位的倾角，确定所述空闲车位符合预先确定的车位形状规则。Determining that the free parking space conforms to a predetermined parking shape rule according to the positions of the two adjacent vertices of the free parking space relative to the vehicle; or And the inclination of the free parking space, it is determined that the free parking space conforms to a predetermined parking space shape rule.

可选地，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置，确定所述空闲车位符合预先确定的车位形状规则，包括：Optionally, according to the positions of two adjacent vertices of the free parking space relative to the vehicle, determining that the free parking space conforms to a predetermined shape rule of the parking space includes:

根据所述空闲车位的两个相邻顶点分别相对于所述车辆的位置，计算所述空闲车位中两个相邻顶点之间的距离；Calculating a distance between two adjacent vertices in the idle parking space according to the positions of the two adjacent vertices of the idle parking space relative to the vehicle;

若所述距离落入预先确定的车位长度集合或车位宽度集合中，则确定所述空闲车位符合预先确定的车位形状规则。If the distance falls into a predetermined parking space length set or parking space width set, it is determined that the free parking space conforms to a predetermined parking space shape rule.

可选地，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置和所述空闲车位的倾角，确定所述空闲车位符合预先确定的车位形状规则，包括：Optionally, determining that the free parking space conforms to a predetermined parking space shape rule according to the position of two adjacent vertices of the free parking space with respect to the vehicle and the inclination of the free parking space, respectively, includes:

根据所述空闲车位的两个相邻顶点分别相对于所述车辆的位置，计算所述空闲车位中两个相邻顶点所连直线的倾角；Calculating the inclination of a straight line connected to the two adjacent vertices in the free parking space according to the positions of the two adjacent vertices in the free parking space relative to the vehicle;

计算所述直线的倾角与所述空闲车位的倾角之间的差值；Calculating a difference between the inclination of the straight line and the inclination of the free parking space;

若所述差值与预设差值之间的误差在预设范围内，则确定所述空闲车位符合预先确定的车位形状规则。If the error between the difference and the preset difference is within a preset range, it is determined that the idle parking space conforms to a predetermined parking space shape rule.

在一种可能的实施方式下，还包括：In a possible implementation manner, the method further includes:

在进行泊车的过程中，循环执行以下操作：During parking, the following operations are performed cyclically:

计算所述目标车位两个相邻顶点分别相对于所述车辆的位置；Calculating the positions of two adjacent vertices of the target parking space relative to the vehicle;

根据所述目标车位两个相邻顶点分别相对于所述车辆的位置，对泊车方位进行调整；Adjusting the parking position according to the positions of two adjacent vertices of the target parking space relative to the vehicle;

直至确定泊车方位与所述目标车位的倾角之间的误差小于预设误差。Until it is determined that the error between the parking orientation and the inclination of the target parking space is less than a preset error.

第二方面，本申请实施例提供的一种自动泊车装置，包括：In a second aspect, an automatic parking device provided by an embodiment of the present application includes:

变换模块，用于接收泊车指令，获取摄像头采集到的周围图像，对所述图像进行透视变换；A transformation module, configured to receive a parking instruction, obtain a surrounding image collected by a camera, and perform perspective transformation on the image;

识别模块，用于将进行透视变换后的图像输入到预先训练的空闲车位特征模型中，利用所述空闲车位特征模型识别出所述图像中符合空闲车位特征的区域，并确定每一区域的倾角和在所述图像中的位置信息，其中，每一区域代表一个空闲车位，区域的倾角为空闲车位的倾角，空闲车位的倾角是指选定的车位中心线与基准线之间的夹角；A recognition module, configured to input a perspective-transformed image into a pre-trained free parking space feature model, use the free parking space feature model to identify areas in the image that conform to free parking space features, and determine the inclination of each area And position information in the image, wherein each area represents an idle parking space, the inclination of the area is the inclination of the idle parking space, and the inclination of the idle parking space refers to the included angle between the centerline of the selected parking space and the reference line;

泊车模块，用于接收目标车位的选择指令，根据所述目标车位的倾角和相对于所述车辆的位置为所述车辆规划泊车线路，并根据规划的泊车线路驱动所述车辆进行泊车。A parking module for receiving a selection instruction of a target parking space, planning a parking line for the vehicle according to the inclination of the target parking space and a position relative to the vehicle, and driving the vehicle to park according to the planned parking line. car.

在一种可能的实施方式下，还包括，In a possible implementation manner, the method further includes:

确定模块，用于在确定所述空闲车位相对于所述车辆的位置之前，确定所述空闲车位符合预先确定的车位形状规则。A determining module, configured to determine that the free parking space conforms to a predetermined parking space shape rule before determining the position of the free parking space relative to the vehicle.

在一种可能的实施方式下，所述确定模块具体用于：In a possible implementation manner, the determining module is specifically configured to:

根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置，确定所述空闲车位符合预先确定的车位形状规则；或者，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置和所述空闲车位的倾角，确定所述空闲车位符合预先确定的车位形状规则。Determining that the free parking space conforms to a predetermined parking shape rule according to the positions of the two adjacent vertices of the free parking space relative to the vehicle; or, based on the two adjacent vertices of the free parking space, respectively, relative to the vehicle And the inclination of the free parking space, it is determined that the free parking space conforms to a predetermined parking space shape rule.

调整模块，用于在进行泊车的过程中，循环执行以下操作：计算所述目标车位两个相邻顶点分别相对于所述车辆的位置；根据所述目标车位两个相邻顶点分别相对于所述车辆的位置，对泊车方位进行调整；直至确定泊车方位与所述目标车位的倾角之间的误差小于预设误差。An adjustment module is configured to perform the following operations in a cyclic manner during parking: calculate the positions of two adjacent vertices of the target parking space relative to the vehicle; The position of the vehicle is adjusted for the parking orientation; until the error between the parking orientation and the inclination of the target parking space is determined to be less than a preset error.

第三方面，本申请实施例提供的一种电子设备，包括：至少一个处理器，以及与所述至少一个处理器通信连接的存储器，其中：According to a third aspect, an electronic device provided in an embodiment of the present application includes: at least one processor, and a memory communicatively connected to the at least one processor, wherein:

存储器存储有可被至少一个处理器执行的指令，该指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行上述自动泊车方法。The memory stores instructions executable by at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the automatic parking method described above.

第四方面，本申请实施例提供的一种计算机可读介质，存储有计算机可执行指令，所述计算机可执行指令用于执行上述自动泊车方法。In a fourth aspect, a computer-readable medium provided by an embodiment of the present application stores computer-executable instructions, where the computer-executable instructions are used to execute the foregoing automatic parking method.

另外，第二方面至第四方面中任一种设计方式所带来的技术效果可参见第一方面中不同实现方式所带来的技术效果，此处不再赘述。In addition, for the technical effects brought by any one of the design methods in the second aspect to the fourth aspect, refer to the technical effects brought by the different implementation manners in the first aspect, which will not be repeated here.

本申请的这些方面或其它方面在以下实施例的描述中会更加简明易懂。These or other aspects of this application will be more concise and easy to understand in the description of the following embodiments.

附图说明BRIEF DESCRIPTION OF THE DRAWINGS

此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：The drawings described here are used to provide a further understanding of the present application and constitute a part of the present application. The schematic embodiments of the present application and the description thereof are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

图1为现有技术提供的停车位的示意图；1 is a schematic diagram of a parking space provided in the prior art;

图2a、图2b、图2c为本申请实施例提供不同倾角的空闲车位的示意图；2a, 2b, and 2c are schematic diagrams of free parking spaces with different inclination angles according to an embodiment of the present application;

图3为本申请实施例提供的利用鱼眼摄像头拍摄到的车位的图像；3 is an image of a parking space captured by a fisheye camera according to an embodiment of the present application;

图4为本申请实施例提供的对图3中的图像进行透视变换后的图像；4 is an image obtained by performing perspective transformation on the image in FIG. 3 according to an embodiment of the present application;

图5为本申请实施例提供的自动泊车的应用场景示意图；5 is a schematic diagram of an application scenario of automatic parking provided by an embodiment of the present application;

图6为本申请实施例提供的对空闲车位进行标注的示意图；FIG. 6 is a schematic diagram of marking a free parking space provided by an embodiment of the present application; FIG.

图7为本申请实施例提供的对空闲车位进行标注后得到的数据的示意图；FIG. 7 is a schematic diagram of data obtained after an idle parking space is marked according to an embodiment of the present application; FIG.

图8为本申请实施例提供的自动泊车方法的流程图；8 is a flowchart of an automatic parking method according to an embodiment of the present application;

图9为本申请实施例提供的用于实现自动泊车方法的电子设备的硬件结构示意图；9 is a schematic diagram of a hardware structure of an electronic device for implementing an automatic parking method according to an embodiment of the present application;

图10为本申请实施例提供的自动泊车装置的结构示意图。FIG. 10 is a schematic structural diagram of an automatic parking device according to an embodiment of the present application.

具体实施方式detailed description

为了解决现有技术中雷达***不能探测到附近没有障碍物的空闲车位而无法进行自动泊车的问题，本申请实施例提供了一种自动泊车方法及装置。In order to solve the problem that the radar system cannot detect an idle parking space without obstacles in the prior art and cannot perform automatic parking, embodiments of the present application provide an automatic parking method and device.

以下结合说明书附图对本申请的优选实施例进行说明，应当理解，此处所描述的优选实施例仅用于说明和解释本申请，并不用于限定本申请，并且在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。The following describes preferred embodiments of the present application with reference to the accompanying drawings of the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application. In the case of no conflict, the present application The embodiments and features in the embodiments can be combined with each other.

为了便于理解本申请，本申请涉及的技术术语中：In order to facilitate the understanding of this application, the technical terms involved in this application:

泊车，即停车。Parking, that is, parking.

空闲车位的倾角，与车位的倾角含义相同，是指选定的车位中心线与基准线之间的夹角，比如，选定的车位中心线为组成车位的平行四边形中两个长边的中心线，基准线为水平线，那么，图2a中车位的倾角为0°，图2b中车位的倾角为90°，图2c中车位的倾角为135°。The inclination of a parking space has the same meaning as the inclination of a parking space. It refers to the angle between the centerline of the selected parking space and the reference line. For example, the centerline of the selected parking space is the center of the two long sides of the parallelogram that forms the parking space. Line, the baseline is the horizontal line, then the inclination of the parking space in FIG. 2a is 0 °, the inclination of the parking space in FIG. 2b is 90 °, and the inclination of the parking space in FIG. 2c is 135 °.

透视变换，是指将摄像头视图转换到从上到下的俯视视图的过程，也叫鸟瞰图变换，得到的图像也叫鸟瞰图，一般地，如果对物体进行俯视拍摄，物体的形变程度会比较小，而如果从其他角度进行拍摄，物体的形变程度都会比较大，在利用车辆上的摄像头拍摄车位的图像时，明显是无法做到俯视拍摄的，因此，可对摄像头拍摄到的图像进行透视变换，即转换车位的拍摄视角，尽可能地降低图像中车位形状的畸变，如图3所示，为利用鱼眼摄像头拍摄到的车位的图像，图4为对其进行透视变换后的图像。Perspective transformation refers to the process of converting the camera view to a top-down view from top to bottom. It is also called a bird's-eye view transformation. The resulting image is also called a bird's-eye view. Generally, if an object is photographed from above, the degree of deformation of the object will be compared. Small, and if you shoot from other angles, the deformation of the object will be relatively large. When you use the camera on the vehicle to capture the image of the parking space, it is obviously impossible to shoot from the top. Therefore, you can perspective the image captured by the camera. Transform is to change the shooting angle of the parking space to reduce the distortion of the shape of the parking space in the image as much as possible. As shown in FIG. 3, it is an image of a parking space captured by a fisheye camera, and FIG. 4 is an image after perspective transformation.

现有技术中，雷达***在寻找空闲车位时需要依赖空闲车位附近的障碍物，因此，无法发现附近没有障碍物的空闲车位，进而完成自动泊车，因此，发明人想到不依赖障碍物，而借助于视觉图像的方法来识别空闲车位，下面对本申请的技术方案进行详细介绍。In the prior art, the radar system needs to rely on obstacles near the free parking space when searching for free parking spaces. Therefore, it is impossible to find free parking spaces without obstacles nearby, and then complete automatic parking. Therefore, the inventor thought of not relying on obstacles, and The method of visual images is used to identify idle parking spaces. The technical solution of the present application is described in detail below.

如图5所示，为本申请实施例提供的自动泊车的应用场景示意图，当车辆驶入图5所示的停车场、且开启了自动泊车功能，则可对安装在车辆上的摄像头采集到的周围图像，进行透视变换，尽可能地减小车位的畸变程度，之后，将进行透视变换后的图像输入到空闲车位特征模型中，利用该模型识别出图像中符合空闲车位特征的区域，并确定每一区域的倾角和在图像中的位置信息，其中，每一区域代表一个空闲车位，区域的倾角为空闲车位的倾角，之后，针对每一空闲车位，可根据预先标定的摄像头采集的图像中每个像素点和车辆之间的相对位置关系和代表该空闲车位的区域在图像中的位置，确定该空闲车位的相对车位的位置，之后，可从各空闲车位中选择一个作为目标车位，根据目标车位的倾角和相对车辆的位置为车辆规划泊车线路，进而根据规划的泊车线路驱动车辆进行自动泊车。As shown in FIG. 5, it is a schematic diagram of an application scenario of automatic parking provided by an embodiment of the present application. When a vehicle enters the parking lot shown in FIG. 5 and the automatic parking function is enabled, a camera installed on the vehicle can be used. The collected surrounding images are subjected to perspective transformation to minimize the distortion of the parking space. After that, the perspective transformed image is input into the free parking space feature model, and the model is used to identify areas in the image that meet the characteristics of the free parking space. And determine the inclination of each area and the position information in the image, where each area represents an idle parking space, and the inclination of the area is the inclination of the idle parking space. After that, for each idle parking space, it can be collected according to a pre-calibrated camera The relative position relationship between each pixel in the image and the vehicle and the position of the area representing the free parking space in the image. Determine the relative parking position of the free parking space. After that, you can choose one of the free parking spaces as the target. Parking space, according to the inclination of the target parking space and the relative vehicle position, plan the parking line for the vehicle, and then according to the plan Line driver parking the vehicle automatic parking.

其中，空闲车位特征模型是利用深度学习框架模型对预设数量的图像中空闲车位所在区域的图像特征进行学习得到的，空闲车位所在区域的图像特征如图像中空闲车位区域的颜色特征、纹理特征、形状特征和空间关系特征等。Among them, the free parking space feature model is obtained by using a deep learning framework model to learn the image features of the area where the free parking spaces are located in a preset number of images. , Shape features and spatial relationship features.

由于本申请实施例中的自动泊车方法属于一种视觉泊车方法，可完全脱离对障碍物的依赖，因此，能很好地应用于附近没有障碍物的空闲车位，并且，具有更广的应用前景和商业价值。Since the automatic parking method in the embodiment of the present application belongs to a visual parking method, it can completely get rid of the dependence on obstacles, so it can be well applied to idle parking spaces without obstacles nearby, and has a wider Application prospects and business value.

下面结合具体的实施例对上述过程进行详细描述。The above process is described in detail below with reference to specific embodiments.

具体实施时，可以先从停车场（包含各种各样倾角的车位）采集车位图像，一般地，采集到的车位图像都是部分的，为了可以更好地识别出空闲车位的区域特征，可挑选包含空闲车位区域较多的图像进行学习，比如，挑选空闲车位区域大于整个空闲车位区域三分之一的图像，之后，对挑选的每幅图像均进行透视变换，然后对透视变换后图像中的空闲车位进行标注，如图6所示，为对空闲车位进行标注的示意图，其主要标注两类信息，一类是车位两个相邻顶点（图6中的白色点），另一类是车位的倾角，非空闲车位不进行标注，标注完成后可生成如图7所示的txt标注信息，并可对数据作归一化处理，之后，可将数据导入Darknet框架模型中进行深度学习，得到空闲车位特征模型。In specific implementation, you can first collect parking space images from parking lots (including parking spaces with various inclination angles). Generally, the collected parking space images are partial. In order to better identify the regional characteristics of free parking spaces, you can Select images that contain more free parking space areas for learning, for example, select images with free parking space areas larger than one-third of the entire free parking area area, and then perform a perspective transformation on each selected image, and then transform the perspective transformed images into The free parking spaces are labeled. As shown in Figure 6, it is a schematic diagram for labeling free parking spaces. It mainly labels two types of information, one is the two adjacent vertices of the parking space (white dots in Figure 6), and the other is The inclination of the parking space is not marked. Non-idle parking spaces are not labeled. After the labeling is completed, the txt label information shown in Figure 7 can be generated, and the data can be normalized. After that, the data can be imported into the Darknet framework model for deep learning. Get the free parking space feature model.

在图7中，第一列表示类别信息，其中，0代表倾角，1代表顶点，后面依次为空闲车位中心处的（x，y）坐标，空闲车位的宽、高和倾角信息。 In FIG. 7, the first column represents category information, where 0 represents the inclination angle, 1 represents the vertex, followed by (x, y) coordinates at the center of the free parking space, and the width, height, and inclination information of the free parking space.

并且，可利用图像中包含的非空闲车位的图像特征，和/或，未参与学习的图像中包含的空闲车位的图像特征，对得到的空闲车位特征模型进行测试，根据测试结果确定空闲车位特征模型的精度，若精度达不到要求，还可添加训练样本继续进行训练，直到最终得到的空闲车位特征模型的精度符合要求。In addition, the image features of the non-idle parking spaces contained in the images and / or the image features of the free parking spaces contained in the images not participating in the learning may be used to test the obtained free parking space feature models, and the free parking space features may be determined according to the test results The accuracy of the model. If the accuracy cannot meet the requirements, training samples can be added to continue the training until the accuracy of the free parking space feature model finally meets the requirements.

这里，通过大量样本的学***行车位、倾斜车位、半封闭车位、全封闭车位等等。Here, through the learning and training of a large number of samples, the accuracy of the free parking space feature model is guaranteed, and the inclination of the free parking space is additionally considered when training the free parking space feature model, so that the solution provided by the embodiment of the present application can be better adapted Various parking spaces, vertical parking spaces, parallel parking spaces, inclined parking spaces, semi-closed parking spaces, fully enclosed parking spaces, etc.

当接收到自动泊车指令时，即可对摄像头采集到的周围图像进行透视变换，并将进行透视变换后的图像输入到空闲车位特征模型中，识别出图像中符合空闲车位特征的区域，并确定每一区域的倾角和在图像中的位置信息，其中，每一区域代表一个空闲车位，并且，可根据预先确定的车位形状规则对各空闲车位进行筛选。When an automatic parking instruction is received, the surrounding images collected by the camera can be perspective-transformed, and the perspective-transformed image can be input into the free parking space feature model to identify the areas in the image that meet the characteristics of the free parking space, and Determine the inclination of each area and the position information in the image, where each area represents a free parking space, and each free parking space can be filtered according to a predetermined parking space shape rule.

具体地，针对每一空闲车位，可选取图像中代表所述空闲车位两个相邻顶点的像素点，根据预先标定的每一像素点相对于车辆的位置，确定空闲车位两个相邻顶点分别相对于车辆的位置，进而根据空闲车位两个相邻顶点分别相对于车辆的位置，确定该空闲车位符合预先确定的车位特点；或者，根据该空闲车位两个相邻顶点分别相对于车辆的位置和该空闲车位的倾角，确定空闲车位符合预先确定的车位特点。Specifically, for each vacant parking space, pixel points representing two adjacent vertices of the vacant parking space may be selected, and two adjacent vertices of the vacant parking space may be determined respectively according to the position of each pixel relative to the vehicle that is pre-calibrated. Relative to the position of the vehicle, and then according to the positions of the two adjacent vertices of the free parking space relative to the vehicle, it is determined that the free parking space conforms to the characteristics of the predetermined parking space; or, the positions of the two adjacent vertices of the free parking space relative to the vehicle are determined And the inclination of the free parking space, it is determined that the free parking space conforms to the characteristics of a predetermined parking space.

其中，图像中代表空闲车位顶点的可以为一小块区域内的像素点，在具体实施时，可从该区域中选择预设数量的像素点，之后，确定每一像素点相对于车辆的位置，进而对位置数据求平均值，将平均值作为该顶点相对于车辆的位置；也可以，计算选择的预设数量的像素点的坐标平均值，确定与坐标平均值最接近的像素点，将该像素点相对于车辆的位置确定为该顶点相对于车辆的位置，当然，这些都是举例，并不构成对本申请中确定空闲车位顶点相对于车辆的位置的限定。Among them, the apex of the free parking space in the image may be pixels in a small area. In specific implementation, a preset number of pixels may be selected from the area, and then the position of each pixel relative to the vehicle is determined. , And then average the position data, and use the average value as the position of the vertex relative to the vehicle; or, calculate the average value of the coordinates of the selected preset number of pixels, determine the pixel closest to the coordinate average, and The position of the pixel relative to the vehicle is determined as the position of the vertex relative to the vehicle. Of course, these are examples, and do not constitute a limitation on determining the position of the vertex of the free parking space relative to the vehicle in the present application.

可选地，根据空闲车位两个相邻顶点分别相对于车辆的位置，确定空闲车位符合预先确定的车位形状规则的方式为：Optionally, according to the positions of two adjacent vertices of the free parking space relative to the vehicle, it is determined that the free parking space conforms to a predetermined shape rule of the parking space:

根据空闲车位的两个相邻顶点分别相对于车辆的位置，计算空闲车位中两个相邻顶点之间的距离，判断该距离是否落入预先确定的车位长度集合或车位宽度集合中，若是，则确定该车位符合车位特点；否则，确定该车位不符合车位特点，这里，若选定的两个相邻顶点为车位宽度方向上的两个点，如图6中所示的白色点，则可判断距离是否落入预先确定的车位宽度集合中；相应地，若选定的两个相邻顶点为车位长度方向上的两个点，则可判断距离是否落入预先确定的车位长度集合中。Calculate the distance between the two adjacent vertices in the free parking space according to the position of the two adjacent vertices in the free parking space relative to the vehicle, and determine whether the distance falls into the predetermined set of parking space lengths or parking space widths. If so, It is determined that the parking space conforms to the characteristics of the parking space; otherwise, it is determined that the parking space does not conform to the characteristics of the parking space. Here, if the two adjacent vertices selected are two points in the width direction of the parking space, as shown by the white points in FIG. 6, Can determine whether the distance falls into the predetermined set of parking space widths; accordingly, if the two adjacent vertices selected are two points in the length direction of the parking space, it can be determined whether the distance falls into the predetermined set of parking space lengths .

可选地，根据空闲车位两个相邻顶点分别相对于车辆的位置和空闲车位的倾角，确定空闲车位符合预先确定的车位形状规则的方式为：Optionally, according to the position of the two adjacent vertices of the free parking space relative to the vehicle and the inclination of the free parking space, the way to determine that the free parking space conforms to the predetermined shape rules of the parking space is:

根据空闲车位的两个相邻顶点分别相对于车辆的位置，，计算空闲车位中两个相邻顶点所连直线的倾角，计算直线的倾角与该车位的倾角之间的差值，判断该差值与预设差值之间的误差是否在预设范围内，若是，则确定该车位符合车位特点；否则，可确定该车位不符合车位特点，比如，若选定的两个相邻顶点为车位宽度方向上的两个点，对应图6中所示的白色点，则预设差值可为90°；若选定的两个相邻顶点为车位长度方向上的两个点，则预设差值可为0°。According to the positions of the two adjacent vertices of the free parking space relative to the vehicle, calculate the inclination of the straight line connected to the two adjacent vertices in the free parking space, calculate the difference between the inclination of the straight line and the inclination of the parking space, and judge the difference Whether the error between the value and the preset difference is within the preset range. If so, determine that the parking space meets the characteristics of the parking space; otherwise, determine that the parking space does not meet the characteristics of the parking space. For example, if the two adjacent vertices selected are The two points in the width direction of the parking space correspond to the white points shown in Figure 6. The preset difference can be 90 °; if the two adjacent vertices selected are two points in the length direction of the parking space, the The difference can be set to 0 °.

进一步地，针对确定出的符合车位形状规则的每一空闲车位，可根据预先标定的摄像头采集到的图像中每一像素点相对于车辆的位置关系和该空闲车位在图像中的位置，确定该空闲车位相对于车辆的位置，之后，可从确定出位置和倾角的各空闲车位中选择一个车位作为目标车位，根据目标车位的倾角和其相对于车位的位置为车辆规划泊车线路，进而根据规划的泊车线路驱动车辆进行泊车。Further, for each determined free parking space that conforms to the parking space shape rule, the position relationship of each pixel in the image collected by a pre-calibrated camera relative to the vehicle and the position of the free parking space in the image may be determined. The position of the free parking space relative to the vehicle. After that, one parking space can be selected as the target parking space from each of the free parking spaces whose position and inclination are determined. According to the inclination of the target parking space and its position relative to the parking space, a parking line is planned for the vehicle, and then according to The planned parking line drives vehicles for parking.

并且，在进行泊车的过程中，还可利用车辆后方的摄像头跟踪目标车位的两个顶点，实时地对泊车时的方位进行微调。In addition, during the parking process, the camera at the rear of the vehicle can also be used to track the two vertices of the target parking space to fine-tune the orientation during parking in real time.

具体地，可在进行泊车的过程中，循环执行以下操作：计算目标车位两个相邻顶点分别相对于车辆的位置，根据目标车位两个相邻顶点分别相对于车辆的位置，对泊车方位进行调整，直至确定泊车方位与目标车位的倾角之间的误差小于预设误差。Specifically, during the parking process, the following operations can be performed in a loop: Calculate the positions of the two adjacent vertices of the target parking space relative to the vehicle, and calculate the positions of the two adjacent vertices of the target parking space relative to the vehicle. The orientation is adjusted until the error between the determined parking orientation and the inclination of the target parking space is less than a preset error.

在本申请实施例中，空闲车位特征模型具有较强的鲁棒性及适应性，可识别出不同倾角和形状的空闲车位，对常见的地下车库具有很好的使用效果，不需要根据相邻车辆进行是否为空闲车位的判断，可扩大自动泊车的使用场景，削弱人工泊车的干预，大大提高了车辆的智能程度，并且，在进行泊车时加入了角点实时跟踪模块（主要是泊车过程中，采用后视摄像头进行实时检测及跟踪），可动态评估车辆和车位的相对角度及位置信息，实时反馈给中心计算单元，形成完善的闭环泊车***。In the embodiment of the present application, the free parking space feature model has strong robustness and adaptability, and can identify free parking spaces with different inclination and shapes, which has a good use effect on common underground garages, and does not need to be based on adjacent The determination of whether the vehicle is a free parking space can expand the use of automatic parking, weaken the intervention of manual parking, greatly improve the intelligence of the vehicle, and add a corner real-time tracking module (mainly when During the parking process, the rear-view camera is used for real-time detection and tracking). The relative angle and position information of the vehicle and the parking space can be dynamically evaluated and fed back to the central computing unit in real time to form a complete closed-loop parking system.

如图8所示，为本申请实施例提供的自动泊车方法的流程图，包括以下步骤：As shown in FIG. 8, a flowchart of an automatic parking method according to an embodiment of the present application includes the following steps:

S81：接收泊车指令，获取摄像头采集的周围图像，对周围图像进行透视变换。S81: Receive a parking instruction, acquire a surrounding image collected by a camera, and perform perspective transformation on the surrounding image.

S802：将进行透视变换后的图像输入到空闲车位特征模型中，利用空闲车位特征模型识别出图像中符合空闲车位特征的区域，并确定每一区域的倾角和在图像中的位置信息，其中，每一区域代表一个空闲车位。S802: Input the perspective-transformed image into the free parking space feature model, use the free parking space feature model to identify areas in the image that match the free parking space feature, and determine the inclination of each area and the position information in the image, where: Each area represents a free parking space.

其中，空闲车位的倾角是指选定的车位中心线与基准线之间的夹角，空闲车位特征模型是利用深度学习框架模型对预设数量的图像中空闲车位所在区域的图像特征进行学习得到的。Among them, the inclination of the free parking space refers to the included angle between the centerline of the selected parking space and the reference line. The free parking space feature model is obtained by learning the image features of the area where the free parking space is located in a preset number of images using a deep learning frame model. of.

S803：根据预先确定的车位形状规则对识别出的每一空闲车位进行筛选。S803: Screen each identified idle parking space according to a predetermined parking space shape rule.

具体的筛选方式参见前述的两种示例情况，在此不再赘述。For specific screening methods, refer to the foregoing two example cases, and details are not described herein again.

S804：针对筛选出的每一空闲车位，根据预先标定的摄像头采集到的图像中每一像素点相对于车辆的位置关系和代表空闲车位的区域在图像中的位置，确定空闲车位相对于车辆的位置。S804: For each screened free parking space, determine the position of the free parking space relative to the vehicle according to the position relationship of each pixel in the image collected by the pre-calibrated camera relative to the vehicle and the position of the area representing the free parking space in the image. position.

这里，车辆在出厂时，已根据每个摄像头的安装位置对其采集的图像进行了标定，确定了其所拍摄的图像中每个像素点相对于车位的位置，建立了位置关系对应表，所以，在得到摄像头所拍摄的图像中每个空闲车位的位置之后，即可根据该位置关系对应表和空闲车位在图像中的位置，确定出空闲车位相对于车位的位置。Here, when the vehicle leaves the factory, the image collected by each camera has been calibrated according to the installation position of each camera, the position of each pixel in the image taken with respect to the parking space is determined, and a position relationship correspondence table is established, so After obtaining the position of each free parking space in the image captured by the camera, the position of the free parking space relative to the parking space can be determined according to the position relationship correspondence table and the position of the free parking space in the image.

S805：接收目标车位的选择指令，从各空闲车位中选择一个车位作为目标车位。S805: Receive a selection instruction of a target parking space, and select a parking space from each idle parking space as a target parking space.

在具体实施时，利用空闲车位特征模型还可得到图像中包含的每个空闲车位的置信度，所以，从各空闲车位中选择目标车位时，可以先确定各空闲车位中置信度大于预设值的空闲车位，再从置信度大于预设值的空闲车位中选择一个作为目标车位。In specific implementation, the confidence level of each free parking space included in the image can also be obtained by using the free parking space feature model. Therefore, when selecting a target parking space from each free parking space, it can be determined that the confidence level of each free parking space is greater than a preset value. , And then select one of the free parking spaces with a confidence level greater than a preset value as the target parking space.

S806：根据目标车位的倾角和相对于车辆的位置为车辆规划泊车线路。S806: Plan a parking line for the vehicle according to the inclination of the target parking space and the position relative to the vehicle.

S807：根据规划的泊车线路驱动车辆进行自动泊车。S807: Drive the vehicle for automatic parking according to the planned parking line.

这里，在自动泊车的过程中，还可利用摄像头跟踪目标车位的顶点位置，根据目标车位的顶点位置对泊车方位进行微调。Here, in the process of automatic parking, a camera can also be used to track the vertex position of the target parking space, and fine-tune the parking orientation according to the vertex position of the target parking space.

具体地，在进行泊车的过程中，循环执行：计算目标车位两个相邻顶点分别相对于车辆的位置，根据目标车位两个相邻顶点分别相对于车辆的位置，对泊车方位进行调整，直至确定泊车方位与目标车位的倾角之间的误差小于预设误差。Specifically, in the process of parking, the loop execution is performed: calculating the positions of two adjacent vertices of the target parking space relative to the vehicle, and adjusting the parking position according to the positions of the two adjacent vertices of the target parking space relative to the vehicle. Until the error between the determined parking orientation and the inclination of the target parking space is less than a preset error.

参见图9所示，为本申请实施例提供的一种电子设备的结构示意图，该电子设备包括收发器901以及处理器902等物理器件，其中，处理器902可以是一个中央处理单元（central processing unit，CPU）、微处理器、专用集成电路、可编程逻辑电路、大规模集成电路、或者为数字处理单元等等。收发器901用于电子设备和其他设备进行数据收发。9 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device includes physical devices such as a transceiver 901 and a processor 902. The processor 902 may be a central processing unit (central processing unit (CPU), microprocessor, application specific integrated circuit, programmable logic circuit, large scale integrated circuit, or digital processing unit, etc. The transceiver 901 is used for data transmission and reception between the electronic device and other devices.

该电子设备还可以包括存储器903用于存储处理器902执行的软件指令，当然还可以存储电子设备需要的一些其他数据，如电子设备的标识信息、电子设备的加密信息、用户数据等。存储器903可以是易失性存储器（volatile memory），例如随机存取存储器（random-access memory，RAM）；存储器903也可以是非易失性存储器（non-volatile memory），例如只读存储器（read-only memory，ROM），快闪存储器（flash memory），硬盘（hard disk drive，HDD）或固态硬盘（solid-state drive，SSD）、或者存储器903是能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质，但不限于此。存储器903可以是上述存储器的组合。The electronic device may further include a memory 903 for storing software instructions executed by the processor 902, and of course, it may also store some other data required by the electronic device, such as identification information of the electronic device, encrypted information of the electronic device, user data, and the like. The memory 903 may be a volatile memory (volatile memory), such as random-access memory (RAM); memory 903 may also be non-volatile memory (non-volatile memory), such as read-only memory (ROM), flash memory, hard disk drive (HDD) or solid state drive (SSD), or memory 903 is any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 903 may be a combination of the above-mentioned memories.

本申请实施例中不限定上述处理器902、存储器903以及收发器901之间的具体连接介质。本申请实施例在图9中仅以存储器903、处理器902以及收发器901之间通过总线904连接为例进行说明，总线在图9中以粗线表示，其它部件之间的连接方式，仅是进行示意性说明，并不引以为限。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图9中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。A specific connection medium between the processor 902, the memory 903, and the transceiver 901 is not limited in the embodiment of the present application. In the embodiment of the present application, only the memory 903, the processor 902, and the transceiver 901 are connected by using a bus 904 as an example in FIG. 9. The bus is shown by a thick line in FIG. 9. The connection modes of other components are only It is for illustrative purposes and is not limited. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only a thick line is used in FIG. 9, but it does not mean that there is only one bus or one type of bus.

处理器902可以是专用硬件或运行软件的处理器，当处理器902可以运行软件时，处理器902读取存储器903存储的软件指令，并在所述软件指令的驱动下，执行前述实施例中涉及的自动泊车方法。The processor 902 may be dedicated hardware or a processor running software. When the processor 902 can run software, the processor 902 reads the software instructions stored in the memory 903 and is driven by the software instructions to execute the foregoing embodiments. The method of automatic parking involved.

当本申请实施例中提供的方法以软件或硬件或软硬件结合实现的时候，电子设备中可以包括多个功能模块，每个功能模块可以包括软件、硬件或其结合。具体的，参见图10所示，为本申请实施例提供的自动泊车装置的结构示意图，包括变换模块1001、识别模块1002、泊车模块1003。When the method provided in the embodiment of the present application is implemented by software or hardware or a combination of software and hardware, the electronic device may include multiple functional modules, and each functional module may include software, hardware, or a combination thereof. Specifically, referring to FIG. 10, a schematic structural diagram of an automatic parking device according to an embodiment of the present application includes a conversion module 1001, an identification module 1002, and a parking module 1003.

变换模块1001，用于接收泊车指令，获取摄像头采集到的周围图像，对所述图像进行透视变换；A transformation module 1001, configured to receive a parking instruction, obtain a surrounding image collected by a camera, and perform perspective transformation on the image;

识别模块1002，用于将进行透视变换后的图像输入到预先训练的空闲车位特征模型中，利用所述空闲车位特征模型识别出所述图像中符合空闲车位特征的区域，并确定每一区域的倾角和在所述图像中的位置信息，其中，每一区域代表一个空闲车位，区域的倾角为空闲车位的倾角，空闲车位的倾角是指选定的车位中心线与基准线之间的夹角；A recognition module 1002 is configured to input a perspective-transformed image into a pre-trained free parking space feature model, use the free parking space feature model to identify areas in the image that conform to free parking space features, and determine the area of each area. Inclination and position information in the image, where each area represents an idle parking space, the inclination of the area is the inclination of the idle parking space, and the inclination of the idle parking space refers to the angle between the centerline of the selected parking space and the reference line ;

泊车模块1003，用于接收目标车位的选择指令，根据所述目标车位的倾角和相对于所述车辆的位置为所述车辆规划泊车线路，并根据规划的泊车线路驱动所述车辆进行泊车。A parking module 1003 is configured to receive a selection instruction of a target parking space, plan a parking line for the vehicle according to the inclination of the target parking space and a position relative to the vehicle, and drive the vehicle to perform the planning according to the planned parking line. Parking.

确定模块1004，用于在确定所述空闲车位相对于所述车辆的位置之前，确定所述空闲车位符合预先确定的车位形状规则。A determining module 1004 is configured to determine, before determining the position of the free parking space relative to the vehicle, that the free parking space conforms to a predetermined shape rule of the parking space.

在一种可能的实施方式下，所述确定模块1004具体用于：In a possible implementation manner, the determining module 1004 is specifically configured to:

调整模块1005，用于在进行泊车的过程中，循环执行以下操作：计算所述目标车位两个相邻顶点分别相对于所述车辆的位置；根据所述目标车位两个相邻顶点分别相对于所述车辆的位置，对泊车方位进行调整；直至确定泊车方位与所述目标车位的倾角之间的误差小于预设误差。An adjustment module 1005 is configured to perform the following operations in a loop during parking: calculate the positions of two adjacent vertices of the target parking space relative to the vehicle; Adjusting the parking position at the position of the vehicle; until the error between the parking position and the inclination of the target parking space is determined to be less than a preset error.

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。各个模块相互之间的耦合可以是通过一些接口实现，这些接口通常是电性通信接口，但是也不排除可能是机械接口或其它的形式接口。因此，作为分离部件说明的模块可以是或者也可以不是物理上分开的，既可以位于一个地方，也可以分布到同一个或不同设备的不同位置上。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。The division of the modules in the embodiments of the present application is schematic and is only a logical function division. In actual implementation, there may be another division manner. In addition, the functional modules in the embodiments of the present application may be integrated into one process. In the device, it can also exist separately physically, or two or more modules can be integrated into one module. The coupling between the various modules can be achieved through some interfaces. These interfaces are usually electrical communication interfaces, but it is not excluded that they may be mechanical interfaces or other forms of interfaces. Therefore, the modules described as separate components may or may not be physically separated, and may be located in one place or distributed to different locations on the same or different devices. The above integrated modules may be implemented in the form of hardware or software functional modules.

本申请实施例还提供了一种计算机可读存储介质，存储为执行上述处理器所需执行的计算机可执行指令，其包含用于执行上述处理器所需执行的程序。An embodiment of the present application further provides a computer-readable storage medium that stores computer-executable instructions that need to be executed to execute the processor, and includes a program that is required to execute the processor.

在一些可能的实施方式中，本申请提供的自动泊车方法的各个方面还可以实现为一种程序产品的形式，其包括程序代码，当所述程序产品在电子设备上运行时，所述程序代码用于使所述电子设备执行本说明书上述描述的根据本申请各种示例性实施方式的自动泊车方法中的步骤。In some possible implementation manners, various aspects of the automatic parking method provided in the present application may also be implemented in the form of a program product, which includes program code. When the program product runs on an electronic device, the program The code is used to cause the electronic device to perform the steps in the automatic parking method according to various exemplary embodiments of the present application described above in the present specification.

所述程序产品可以采用一个或多个可读介质的任意组合。可读介质可以是可读信号介质或者可读存储介质。可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的***、装置或器件，或者任意以上的组合。可读存储介质的更具体的例子（非穷举的列表）包括：具有一个或多个导线的电连接、便携式盘、硬盘、随机存取存储器（RAM）、只读存储器（ROM）、可擦式可编程只读存储器（EPROM或闪存）、光纤、便携式紧凑盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: electrical connections with one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable Programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.

本申请的实施方式的用于自动泊车的程序产品可以采用便携式紧凑盘只读存储器(CD-ROM)并包括程序代码，并可以在计算设备上运行。然而，本申请的程序产品不限于此，在本文件中，可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行***、装置或者器件使用或者与其结合使用。The program product for automatic parking according to the embodiment of the present application may adopt a portable compact disk read-only memory (CD-ROM) and include a program code, and may run on a computing device. However, the program product of the present application is not limited thereto. In this document, the readable storage medium may be any tangible medium containing or storing a program, and the program may be used by or in combination with an instruction execution system, apparatus, or device.

可读信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了可读程序代码。这种传播的数据信号可以采用多种形式，包括——但不限于——电磁信号、光信号或上述的任意合适的组合。可读信号介质还可以是可读存储介质以外的任何可读介质，该可读介质可以发送、传播或者传输用于由指令执行***、装置或者器件使用或者与其结合使用的程序。The readable signal medium may include a data signal that is borne in baseband or propagated as part of a carrier wave, in which readable program code is carried. Such a propagated data signal may take many forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination of the foregoing. The readable signal medium may also be any readable medium other than a readable storage medium, and the readable medium may send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

可读介质上包含的程序代码可以用任何适当的介质传输，包括——但不限于——无线、有线、光缆、RF等等，或者上述的任意合适的组合。Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

可以以一种或多种程序设计语言的任意组合来编写用于执行本申请操作的程序代码，所述程序设计语言包括面向对象的程序设计语言—诸如Java、C++等，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算设备上执行、部分地在用户设备上执行、作为一个独立的软件包执行、部分在用户计算设备上部分在远程计算设备上执行、或者完全在远程计算设备或服务器上执行。在涉及远程计算设备的情形中，远程计算设备可以通过任意种类的网络——包括局域网（LAN）或广域网（WAN）—连接到用户计算设备，或者，可以连接到外部计算设备（例如利用因特网服务提供商来通过因特网连接）。The program code for performing the operations of this application may be written in any combination of one or more programming languages, which includes object-oriented programming languages—such as Java, C ++, etc., and also includes conventional procedural Programming language—such as "C" or a similar programming language. The program code may be executed entirely on the user computing device, partly on the user device, as an independent software package, partly on the user computing device, partly on the remote computing device, or entirely on the remote computing device or server On. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (such as using Internet services Provider to connect via the Internet).

应当注意，尽管在上文详细描述中提及了装置的若干单元或子单元，但是这种划分仅仅是示例性的并非强制性的。实际上，根据本申请的实施方式，上文描述的两个或更多单元的特征和功能可以在一个单元中具体化。反之，上文描述的一个单元的特征和功能可以进一步划分为由多个单元来具体化。It should be noted that although several units or sub-units of the device are mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of the two or more units described above may be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided into multiple units to be embodied.

此外，尽管在附图中以特定顺序描述了本申请方法的操作，但是，这并非要求或者暗示必须按照该特定顺序来执行这些操作，或是必须执行全部所示的操作才能实现期望的结果。附加地或备选地，可以省略某些步骤，将多个步骤合并为一个步骤执行，和/或将一个步骤分解为多个步骤执行。Furthermore, although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be split into multiple steps for execution.

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 a method, a system, or a computer program product. Therefore, this application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

本申请是参照根据本申请实施例的方法、装置（***）、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。This application is described with reference to the flowcharts and / or block diagrams of the method, apparatus (system), and computer program product according to the embodiments of the present application. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and combinations of processes and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, so that the instructions generated by the processor of the computer or other programmable data processing device are used to generate instructions Means for realizing the functions specified in one or more flowcharts and / or one or more blocks of the block diagrams.

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 device to work in a particular manner such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, the instructions The device implements the functions specified in one or more flowcharts 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 on a computer or other programmable data processing device, so that a series of steps can be performed on the computer or other programmable device to produce a computer-implemented process, which can be executed on the computer or other programmable device. The instructions provide steps for implementing the functions specified in one or more flowcharts and / or one or more blocks of the block diagrams.

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。Although the preferred embodiments of the present application have been described, those skilled in the art can make other changes and modifications to these embodiments once they know the basic inventive concepts. Therefore, the following claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of this application.

显然，本领域的技术人员可以对本申请进行各种改动和变型而不脱离本申请的精神和范围。这样，倘若本申请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。Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. In this way, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these changes and variations.

Claims

一种自动泊车方法，其特征在于，包括： An automatic parking method, comprising:

接收泊车指令，获取摄像头采集到的周围图像，对所述图像进行透视变换；Receiving a parking instruction, obtaining a surrounding image collected by a camera, and performing perspective transformation on the image;

将进行透视变换后的图像输入到预先训练的空闲车位特征模型中，利用所述空闲车位特征模型识别出所述图像中符合空闲车位特征的区域，并确定每一区域的倾角和在所述图像中的位置信息，其中，每一区域代表一个空闲车位，区域的倾角为空闲车位的倾角，空闲车位的倾角是指选定的车位中心线与基准线之间的夹角；The perspective-transformed image is input into a pre-trained free parking space feature model, and the free parking space feature model is used to identify areas in the image that conform to the free parking space feature, and determine the inclination of each area and the image in the image. Position information in the system, where each area represents an idle parking space, the inclination of the area is the inclination of the idle parking space, and the inclination of the idle parking space refers to the angle between the selected parking space centerline and the reference line;

针对每一空闲车位，根据预先标定的所述摄像头采集到的图像中每一像素点相对于所述车辆的位置关系和代表所述空闲车位的区域在所述图像中的位置，确定所述空闲车位相对于所述车辆的位置； For each vacant parking space, determine the vacancy according to the positional relationship of each pixel in the image collected by the camera pre-calibrated relative to the vehicle and the position of the area representing the vacant parking space in the image. The position of the parking space relative to the vehicle; Ranch

接收目标车位的选择指令，根据所述目标车位的倾角和相对于所述车辆的位置为所述车辆规划泊车线路，并根据规划的泊车线路驱动所述车辆进行泊车。 Receive a selection instruction of a target parking space, plan a parking line for the vehicle according to the inclination of the target parking space and a position relative to the vehicle, and drive the vehicle to park according to the planned parking line. Ranch
如权利要求1所述的方法，其特征在于，确定所述空闲车位相对于所述车辆的位置之前，还包括：The method according to claim 1, before determining the position of the free parking space relative to the vehicle, further comprising:

确定所述空闲车位符合预先确定的车位形状规则。It is determined that the free parking space conforms to a predetermined parking space shape rule.
如权利要求2所述的方法，其特征在于，确定所述空闲车位符合预先确定的车位形状规则，包括：The method according to claim 2, wherein determining that the free parking space conforms to a predetermined parking space shape rule comprises:

选取所述图像中代表所述空闲车位两个相邻顶点的像素点；Selecting pixels in the image that represent two adjacent vertices of the free parking space;

根据预先标定的每一像素点相对于所述车辆的位置，确定所述空闲车位两个相邻顶点分别相对于所述车辆的位置；Determining the positions of two adjacent vertices of the free parking space relative to the vehicle according to the positions of each pixel relative to the vehicle in advance;

根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置，确定所述空闲车位符合预先确定的车位形状规则；或者，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置和所述空闲车位的倾角，确定所述空闲车位符合预先确定的车位形状规则。Determining that the free parking space conforms to a predetermined parking shape rule according to the positions of the two adjacent vertices of the free parking space relative to the vehicle; or, based on the two adjacent vertices of the free parking space, respectively, relative to the vehicle And the inclination of the free parking space, it is determined that the free parking space conforms to a predetermined parking space shape rule.
如权利要求3所述的方法，其特征在于，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置，确定所述空闲车位符合预先确定的车位形状规则，包括：The method according to claim 3, wherein determining that the free parking space conforms to a predetermined parking space shape rule according to the positions of two adjacent vertices of the free parking space relative to the vehicle, respectively, comprising:

根据所述空闲车位的两个相邻顶点分别相对于所述车辆的位置，计算所述空闲车位中两个相邻顶点之间的距离；Calculating a distance between two adjacent vertices in the idle parking space according to the positions of the two adjacent vertices of the idle parking space relative to the vehicle;

若所述距离落入预先确定的车位长度集合或车位宽度集合中，则确定所述空闲车位符合预先确定的车位形状规则。If the distance falls into a predetermined parking space length set or parking space width set, it is determined that the free parking space conforms to a predetermined parking space shape rule.
如权利要求3所述的方法，其特征在于，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置和所述空闲车位的倾角，确定所述空闲车位符合预先确定的车位形状规则，包括：The method according to claim 3, wherein, according to a position of two adjacent vertices of the free parking space with respect to the position of the vehicle and an inclination of the free parking space, determining that the free parking space conforms to a predetermined shape of the parking space Rules, including:

根据所述空闲车位的两个相邻顶点分别相对于所述车辆的位置，计算所述空闲车位中两个相邻顶点所连直线的倾角；Calculating the inclination of a straight line connected to the two adjacent vertices in the free parking space according to the positions of the two adjacent vertices in the free parking space relative to the vehicle;

计算所述直线的倾角与所述空闲车位的倾角之间的差值；Calculating a difference between the inclination of the straight line and the inclination of the free parking space;

若所述差值与预设差值之间的误差在预设范围内，则确定所述空闲车位符合预先确定的车位形状规则。If the error between the difference and the preset difference is within a preset range, it is determined that the idle parking space conforms to a predetermined parking space shape rule.
如权利要求3~5任一所述的方法，其特征在于，还包括：The method according to any one of claims 3 to 5, further comprising:

在进行泊车的过程中，循环执行以下操作：During parking, the following operations are performed cyclically:

计算所述目标车位两个相邻顶点分别相对于所述车辆的位置；Calculating the positions of two adjacent vertices of the target parking space relative to the vehicle;

根据所述目标车位两个相邻顶点分别相对于所述车辆的位置，对泊车方位进行调整；Adjusting the parking position according to the positions of two adjacent vertices of the target parking space relative to the vehicle;

直至确定泊车方位与所述目标车位的倾角之间的误差小于预设误差。Until it is determined that the error between the parking orientation and the inclination of the target parking space is less than a preset error.
一种自动泊车装置，其特征在于，包括：An automatic parking device, comprising:

变换模块，用于接收泊车指令，获取摄像头采集到的周围图像，对所述图像进行透视变换；A transformation module, configured to receive a parking instruction, obtain a surrounding image collected by a camera, and perform perspective transformation on the image;

识别模块，用于将进行透视变换后的图像输入到预先训练的空闲车位特征模型中，利用所述空闲车位特征模型识别出所述图像中符合空闲车位特征的区域，并确定每一区域的倾角和在所述图像中的位置信息，其中，每一区域代表一个空闲车位，区域的倾角为空闲车位的倾角，空闲车位的倾角是指选定的车位中心线与基准线之间的夹角；A recognition module, configured to input a perspective-transformed image into a pre-trained free parking space feature model, use the free parking space feature model to identify areas in the image that conform to free parking space features, and determine the inclination of each area And position information in the image, wherein each area represents an idle parking space, the inclination of the area is the inclination of the idle parking space, and the inclination of the idle parking space refers to the included angle between the centerline of the selected parking space and the reference line;

泊车模块，用于接收目标车位的选择指令，根据所述目标车位的倾角和相对于所述车辆的位置为所述车辆规划泊车线路，并根据规划的泊车线路驱动所述车辆进行泊车。A parking module for receiving a selection instruction of a target parking space, planning a parking line for the vehicle according to the inclination of the target parking space and a position relative to the vehicle, and driving the vehicle to park according to the planned parking line. car.
如权利要求7所述的装置，其特征在于，还包括，The apparatus according to claim 7, further comprising:

确定模块，用于在确定所述空闲车位相对于所述车辆的位置之前，确定所述空闲车位符合预先确定的车位形状规则。A determining module, configured to determine that the free parking space conforms to a predetermined parking space shape rule before determining the position of the free parking space relative to the vehicle.
如权利要求8所述的装置，其特征在于，所述确定模块具体用于：The apparatus according to claim 8, wherein the determining module is specifically configured to:

选取所述图像中代表所述空闲车位两个相邻顶点的像素点；Selecting pixels in the image that represent two adjacent vertices of the free parking space;

根据预先标定的每一像素点相对于所述车辆的位置，确定所述空闲车位两个相邻顶点分别相对于所述车辆的位置；Determining the positions of two adjacent vertices of the free parking space relative to the vehicle according to the positions of each pixel relative to the vehicle in advance;

根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置，确定所述空闲车位符合预先确定的车位形状规则；或者，根据所述空闲车位两个相邻顶点分别相对于所述车辆的位置和所述空闲车位的倾角，确定所述空闲车位符合预先确定的车位形状规则。Determining that the free parking space conforms to a predetermined parking shape rule according to the positions of the two adjacent vertices of the free parking space relative to the vehicle; or, based on the two adjacent vertices of the free parking space, respectively, relative to the vehicle And the inclination of the free parking space, it is determined that the free parking space conforms to a predetermined parking space shape rule.
如权利要求9所述的装置，其特征在于，所述确定模块具体用于：The apparatus according to claim 9, wherein the determining module is specifically configured to:

根据所述空闲车位的两个相邻顶点分别相对于所述车辆的位置，计算所述空闲车位中两个相邻顶点之间的距离；Calculating a distance between two adjacent vertices in the idle parking space according to the positions of the two adjacent vertices of the idle parking space relative to the vehicle;

若所述距离落入预先确定的车位长度集合或车位宽度集合中，则确定所述空闲车位符合预先确定的车位形状规则。If the distance falls into a predetermined parking space length set or parking space width set, it is determined that the free parking space conforms to a predetermined parking space shape rule.
如权利要求9所述的装置，其特征在于，所述确定模块具体用于：The apparatus according to claim 9, wherein the determining module is specifically configured to:

根据所述空闲车位的两个相邻顶点分别相对于所述车辆的位置，计算所述空闲车位中两个相邻顶点所连直线的倾角；Calculating the inclination of a straight line connected to the two adjacent vertices in the free parking space according to the positions of the two adjacent vertices in the free parking space relative to the vehicle;

计算所述直线的倾角与所述空闲车位的倾角之间的差值；Calculating a difference between the inclination of the straight line and the inclination of the free parking space;

若所述差值与预设差值之间的误差在预设范围内，则确定所述空闲车位符合预先确定的车位形状规则。If the error between the difference and the preset difference is within a preset range, it is determined that the idle parking space conforms to a predetermined parking space shape rule.
如权利要求9~11任一所述的装置，其特征在于，还包括：The device according to any one of claims 9 to 11, further comprising:

调整模块，用于在进行泊车的过程中，循环执行以下操作：计算所述目标车位两个相邻顶点分别相对于所述车辆的位置；根据所述目标车位两个相邻顶点分别相对于所述车辆的位置，对泊车方位进行调整；直至确定泊车方位与所述目标车位的倾角之间的误差小于预设误差。An adjustment module is configured to perform the following operations in a cyclic manner during parking: calculate the positions of two adjacent vertices of the target parking space relative to the vehicle; The position of the vehicle is adjusted for the parking orientation; until the error between the parking orientation and the inclination of the target parking space is determined to be less than a preset error.
一种电子设备，其特征在于，包括：至少一个处理器，以及与所述至少一个处理器通信连接的存储器，其中：An electronic device, comprising: at least one processor and a memory communicatively connected to the at least one processor, wherein:

所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行如权利要求1至6任一权利要求所述的方法。The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method as claimed in any one of claims 1 to 6. The method described.
一种计算机可读介质，存储有计算机可执行指令，其特征在于，所述计算机可执行指令用于执行如权利要求1至6任一权利要求所述的方法。A computer-readable medium stores computer-executable instructions, wherein the computer-executable instructions are used to execute a method according to any one of claims 1 to 6.