WO2023164786A1 - Intelligent control system and method for vehicle lamp - Google Patents
Intelligent control system and method for vehicle lamp Download PDFInfo
- Publication number
- WO2023164786A1 WO2023164786A1 PCT/CN2022/000029 CN2022000029W WO2023164786A1 WO 2023164786 A1 WO2023164786 A1 WO 2023164786A1 CN 2022000029 W CN2022000029 W CN 2022000029W WO 2023164786 A1 WO2023164786 A1 WO 2023164786A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- vehicle
- light intensity
- light
- value
- car
- Prior art date
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60Q—ARRANGEMENT OF SIGNALLING OR LIGHTING DEVICES, THE MOUNTING OR SUPPORTING THEREOF OR CIRCUITS THEREFOR, FOR VEHICLES IN GENERAL
- B60Q1/00—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor
- B60Q1/02—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments
- B60Q1/04—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60Q—ARRANGEMENT OF SIGNALLING OR LIGHTING DEVICES, THE MOUNTING OR SUPPORTING THEREOF OR CIRCUITS THEREFOR, FOR VEHICLES IN GENERAL
- B60Q1/00—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor
- B60Q1/02—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments
- B60Q1/04—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights
- B60Q1/06—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights adjustable, e.g. remotely-controlled from inside vehicle
- B60Q1/08—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights adjustable, e.g. remotely-controlled from inside vehicle automatically
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F21—LIGHTING
- F21S—NON-PORTABLE LIGHTING DEVICES; SYSTEMS THEREOF; VEHICLE LIGHTING DEVICES SPECIALLY ADAPTED FOR VEHICLE EXTERIORS
- F21S41/00—Illuminating devices specially adapted for vehicle exteriors, e.g. headlamps
- F21S41/60—Illuminating devices specially adapted for vehicle exteriors, e.g. headlamps characterised by a variable light distribution
Definitions
- the invention relates to the field of intelligent vehicles, in particular, an intelligent control system and method for vehicle lights.
- the purpose of the invention is to realize the use of vehicle lights for safe driving.
- the present invention proposes an intelligent control system and method for vehicle lamps to achieve the above objectives. details as follows:
- An intelligent control system for car lights characterized in that it includes a front-of-car video analysis subsystem, a safety light control logic module, a communication subsystem, and a car light control device; the front-of-car video analysis subsystem generates specified front-of-car information, including Designate the environment information in front of the vehicle or the traffic body information in front of the designated vehicle. After the designated front information passes through the safety light control logic module, it generates a signal to control the vehicle light, and connects to the vehicle light control device through the communication subsystem to control the vehicle light;
- the specified environmental information in front of the vehicle includes the specified environmental light intensity category: daytime when the lights do not need to be turned on/very dark when the high beam lights need to be turned on/strong shade when the lights need to be turned on;
- the specified traffic body information in front of the vehicle includes traffic Body category: vehicle/pedestrian including bicycle/vehicle high beam.
- the communication subsystem refers to a wireless or/or wired communication subsystem, including CAN or/or with Bluetooth or/or with WIFI or/or with ModBus or/or with ZigBee.
- the safety light control logic module is embedded in the image processing subsystem or in the communication subsystem or in the vehicle light control device, and is realized by hardware or/or software.
- control vehicle light signal includes turning on or off or flashing the following vehicle lights and combinations thereof: low beam, high beam, and width indicator.
- the vehicle light intelligent control system includes the original manual control vehicle light priority: temporary priority and temporary cut-off; temporary priority means that as long as the human operation turns on the vehicle light, it will enter the original manual control vehicle light state , after the operation is completed and the lights are turned off, the control is automatically restored; temporary cut-off refers to the need to manually operate another cut-off button and restore or restart button to cut off and restore the control; both are implemented with hardware circuits or/or with embedded software .
- An intelligent control method for vehicle lights characterized by comprising: S1, configuring basic parameters for detection; S2, acquiring video images in front of the vehicle; S3, extracting specified information in front of the vehicle from the video images in front of the vehicle, including environmental information; S4, calculating and judging the characteristics of the designated front information of the vehicle: calculating the light intensity of the designated environment information and judging its light intensity level category; S5, the safety light control logic generates a control vehicle light signal according to the specified environment light intensity level category;
- the configuration detection basic parameters include calibrating the internal and external parameters of the camera that provides the video in front of the car, and setting the judgment value of the ambient light intensity level category to determine whether the ambient light intensity category is daytime/very dark/strong shade.
- the setting of the ambient light intensity level category is divided into: 1. During the day, the light intensity that does not need to turn on the headlights; 2. Extremely dark, including the light intensity that needs to be turned on when there are no street lights or weak street lights at night. ;3. Strong shade, the light intensity is stronger than extremely dark, including dusk/early morning/night strong street lights/very low clouds overcast/light intensity that needs to be turned on in tunnels; The gray value is used as its judgment value; or the maximum value of the difference between the gray values of the category at the specified time interval is also set as the judgment value for judging whether it belongs to the difference in ambient light intensity change. , called the environment difference judgment value.
- the method for extracting the specified environment information refers to any algorithm for extracting the target background of the image, including directly extracting the gray value of all pixels in the image, or extracting the gray value of all pixels in the specified environment area in the image.
- the calculating the light intensity of the specified environment information includes calculating a value related to the extracted specified environment information, that is, the gray value, including its average value, as a detection value of the feature of the environment light intensity.
- the judging the specified ambient light intensity level category is to compare the specified ambient light intensity detection value with the set ambient light intensity level category determination value, and the rules include: when specifying the ambient light intensity detection value, ( 1) When it is not less than the daytime judgment value of the set ambient light intensity level category, the ambient light intensity is judged as daytime; (2) When it is not greater than the corresponding extremely dark judgment value, the ambient light intensity is judged as extremely dark; (3) between When the extremely dark judgment value is between the daytime judgment value, the ambient light intensity is judged as strong shade.
- the judging the specified ambient light intensity level category is to compare the specified ambient light intensity detection value with the set ambient light intensity level category judgment value, and refer to the previous image-related light intensity determination and its correlation with the specified ambient light intensity level category.
- the difference between the previous image-related light intensity gray value, its rule includes judging before the above-mentioned rule (1): when the specified ambient light intensity detection value, (0) is greater than its previous image-related light intensity category, it is different from the previous image If the difference between the relevant light intensity gray values is not less than the environment difference judgment value, the environment light intensity is judged as temporary light intensity; otherwise, the above rule (1) is executed.
- the safety light control logic includes:
- S1a configuring the detection basic parameters in S1 also includes setting a traffic body type judgment value, to judge whether the traffic body type is vehicle/pedestrian/high beam
- S3a The specified front-of-vehicle information extracted from the front-of-vehicle video in S3 also includes traffic body information
- S4a calculating and judging the features of the specified front-of-vehicle information in S4 also includes calculating the feature value of the specified traffic body information and judging its category
- S5a the safety light control logic in S5 further includes generating a signal to control the vehicle light according to the category of the designated traffic body in front of the vehicle.
- the set traffic body category judgment value is a set of characteristic values for estimating the smallest identifiable traffic body category, including characteristic width/diameter judgment value, characteristic light intensity level and judgment value of longitudinal span of light zone; 1. Vehicle width: set the conventional minimum vehicle width, and set it several times the pedestrian width; 2. Pedestrian width: set the conventional minimum pedestrian width, and set the motorcycle/motorcycle with the same width; 3.
- Car lights Diameter Set the diameter of the conventional minimum car light, and also set it to be the same as the motorcycle or motorcycle light, and set the light intensity to the minimum light intensity of the car light; Low beam light intensity, the minimum vertical/horizontal span of the light area is set by the conventional maximum vertical/horizontal span of the low beam light area, or the set area is used as the judgment setting of the high beam span; 5. Vehicle low beam light intensity level: Set to be stronger than the light intensity of the car, and the vertical/horizontal span of the light zone is set to the conventional maximum and minimum values; these judgment values can be adjusted and optimized according to the usage.
- the method of extracting the specified traffic body information in front of the vehicle from the video in front of the vehicle includes using the image target detection algorithm to directly extract the contour of the traffic body from the image or from the specified area of the traffic body, and obtain the height, width or diameter of the contour Size and gray value representing its light intensity;
- the image target detection algorithm includes R-CNN algorithm series or YOLO algorithm series, but not limited to these.
- the calculation of the specified traffic body information characteristic value includes extracting the specified traffic body contour width or diameter as its detection value; calculating the average value of the specified traffic body contour light intensity gray value and its longitudinal/transverse direction span as its detection value.
- the determination of the specified traffic body category includes judging vehicles/pedestrians and judging vehicle lights; comparing the detection value of the traffic body outline width or diameter with the set traffic body width judgment value, and conforms to the traffic body judgment rules
- the image target of the image is determined as the corresponding traffic body: vehicle or car light (also a vehicle) or pedestrian or bicycle;
- the set judgment value is comprehensively compared with the set judgment value of the longitudinal/horizontal span, and the image object conforming to the traffic body judgment rules is judged as the corresponding vehicle light: car high beam or car low beam/light.
- the traffic body judging rules for the specified traffic body category are as follows: (1) vehicle and pedestrian judgment: when the detection value of the specified traffic body outline width (1.1) is not less than the traffic body vehicle width setting judgment value, The traffic body is judged to be a vehicle; otherwise, when (1.2) the smaller width detection value is not less than the relevant pedestrian width setting judgment value, it is judged as a pedestrian or bicycle; otherwise, (1.3) or smaller width If the detection value is not less than the judgment value of the diameter/width setting of the relevant car light, and the light intensity detection value is not less than the corresponding car light intensity setting judgment value under strong cloudy/very dark ambient light intensity, then it is judged as a car light , and then judged as a vehicle or a motorcycle; (2) Judgment of vehicle lights: under strong cloudy/very dark ambient light intensity, when the light intensity detection value of the designated traffic body outline area, (2.1) is not less than the traffic body vehicle high beam light intensity Set the judgment value, and at the same time, the detection value of the longitudinal/horizon
- the judgment of the above-mentioned steps (2.1) the high beam of the car also includes the following steps (2.1a): (2.1) is used as a temporary judgment for the judgment of the high beam of the car, and the temporary judgment that has continued several continuous frame images is taken as a far For light judgment, the time required for the number of consecutive frames of images is called the time required for judging the high beam, and the setting is greater than the estimated minimum time for turning on the high beam.
- the safety light control logic also includes:
- the flashing high beam above includes the number of flashes, the interval and the duration of a single flash, and the single duration is shorter than the time required for judging the high beam.
- the present invention can automatically turn on or off the appropriate car lights in time following the changes in the light intensity of the environment and the vehicle lights and pedestrians in front of the car, that is, to avoid forgetting to turn on the car lights in dark light environments and avoid driving at night.
- Accident hazards in vehicles and pedestrians, etc. to ensure safe driving and energy saving; when people think it is necessary to use or not use lights, they can switch independently; at the same time, it is simple, and the video analysis subsystem in front of the car can be embedded in the commonly used driving recorder of the vehicle.
- Figure 1 is a schematic diagram of an intelligent control system for vehicle lights
- Fig. 2 is a flow chart of an intelligent control method for vehicle lights
- Fig. 3 is a schematic diagram of the installation position of the camera lens
- Figure 1 1-1-Video analysis subsystem in front of the car; dotted line indicates the subsystem, including 1-2 to 1-5, and 1-7; 1-2-Video sensor device; 1-3-Front of the car Video data line; 1-4-image processing subsystem, or including safety light control logic such as the rectangular dotted box labeled "light control logic"; 1-5-front information or light control data line; 1-6-communicator System, including 1-7 to 1-9; 1-7-wireless Bluetooth or wired CAN communication device A; 1-8-communication medium, radio wave or wired; 1-9-wireless Bluetooth or wired CAN communication device B terminal, including safety light control logic, such as the rectangular wire frame marked "light control logic”; 1-10-car light control device, including 1-9, and 1-11 to 1-17; 1-11-car light controller, Or include safety light control logic such as marked "light control logic” rectangular dotted line box; 1-12-switch control signal line; 1-13-high beam switch; 1-14-high beam flashing switch; 1-15-near Light switch; 1-16-
- FIG. 1 take a driving video recorder development board equipped with sufficient memory and the ability to calculate and process video images in front of the car as the video analysis subsystem in front of the car.
- the video sensor of 1-2 the front video data line of the module is used as the data line 1-3, and the data processing subsystem centered on the processor module group on the board is used as the image processing subsystem such as 1-4, 1-2
- the acquired front-of-car video image is processed by the image processing subsystem 1-4 through the front-of-car video data line 1-3, and the features of the environment information in front of the car and the traffic body information category in front of the car are extracted.
- the front information data line 1-5 and the communication subsystem 1-6, the safety light control logic embedded in the B terminal generates the car light control signal, which is sent to the car light control device 1-10, wherein the car light controller 1-11 controls the car light Light control signal 1-12 is sent to car light switch 1-13 to 1-16;
- the characteristics of the specified front environment information include: (e1) During the day, it is not necessary to turn on the lights; (e2) It is extremely dark, including no street lights or weak street lights at night, and the high beam lights need to be turned on; (e3) Strong shade, including dusk, In the early morning, strong street lights at night, extremely low clouds and cloudy days, or the lights need to be turned on in the tunnel;
- the specified traffic body information category in front of the vehicle includes (v1) vehicles and pedestrians or bicycles; (v2) the light intensity of the low beam of the vehicle lights is between strong shade and daytime; (v3) the intensity of the high beam lights of the vehicle lights is stronger than that of the daytime;
- the communication subsystem 1-6 includes a bluetooth communication module built in the development board system as the communication subsystem A terminal 1-7, and a wireless microwave 2.4GHz connection with a transmission medium 1-8 placed in the car light control
- the Bluetooth communication module nRF52832 in the device 1-10 is used as the B terminal 1-9 of the communication subsystem, and is transmitted to 1-10;
- the vehicle light control device 1-10 receives the front information data 1-5 sent by the A terminal 1-7 from its Bluetooth communication module B terminal 1-9, and transmits the front information data 1-5 through the safety light control logic.
- the car light controller 1-11 includes a device Darlington-ULN2003 and a solid-state relay SSR module to control 12v to 24v circuit electrical appliances with a 3v electrical signal, or also includes a safety light control logic embedded in the B terminal;
- environmental information characteristics e1 the ambient light intensity is daytime; (e2 ) the ambient light intensity is extremely dark, such as weak street lights or no street lights; (e3) the ambient light intensity is strong shade, such as dusk or early morning or strong street lights at night; (e4) the ambient light intensity is (v2) dipped beam; (e5) ) Ambient light intensity is (v3) high beam; sent to the development board system or
- the car light controller includes a temporary cut-off and a temporary priority circuit for the priority of the original manual car light switch circuit: the car light controller is connected in parallel with the original manual car light switch circuit, and two "normally on” circuits are connected in series in this controller. State"switch; one of them, S of , is used to cut off the controller, and is used as a temporary cut-off; the other, S tof , is controlled by the connection signal S cc of 1-17, which is reversed to the original manual light switch Linkage, that is, when the lights are turned on manually, the controller is cut off, and when the lights are turned off manually, the controller is turned on again, that is, temporary priority.
- S1e configure the detection reference parameters: calibrate the internal and external parameters of the video sensor-camera lens of the intelligent control system for vehicle lights, set the judgment value of the information in front of the vehicle, including setting the judgment value of the environmental light intensity in front of the vehicle and the traffic in front of the vehicle Judgment value of body category;
- the internal and external parameters of the calibration camera include (1) the rearview mirror in the front windshield of the car as shown in Figure 3 marked 3-1 as marked 3-2 back to install the camera, its lens 3-3 points to the front of the car,
- the lateral pull-out line 3-4′′ in the plane where the horizontal line 3-4 is located and the intersection line 3-10 of the car’s longitudinal symmetry plane and the ground are on the same plane.
- H c about 1.5 meters; (2) adjust and fix the camera lens so that the boundary position between the ground and the sky in the camera picture is 3-12, and this embodiment makes the 100-meter line D t from the camera lens and the horizontal center line of the picture 3-9
- this embodiment makes the 100-meter line D t from the camera lens and the horizontal center line of the picture 3-9
- obtain the depression angle ⁇ c between the central axis 3-5 of the camera lens and the horizontal line 3-4 and obtain the above-mentioned dividing line located close to the upper part of the horizontal center line of the camera picture, such as marked 3-12, and the area from below to the lower edge of the image including the road
- There are 6 distances such as f, which can be calibrated before installation; the internal and external parameters
- the environmental light intensity category in front of the car includes daytime, extremely dark, and strong shade, and its light intensity level setting judgment value includes daytime Y ed , extremely dark Y eb , and the environmental difference ⁇ Y ex ( ⁇ T);
- the environmental difference ⁇ Y ex ( ⁇ T) is the maximum judging value of the gray value difference between the specified ambient light intensity categories within the specified time interval ⁇ T, the value smaller than this difference is regarded as the ambient light intensity difference, and the enhancement greater than the difference value is temporary light intensity
- It is represented by the set gray value of the camera video image, and its setting methods include digital setting, imaging setting and automatic setting; the digital setting refers to directly representing daytime Y ed , extremely dark Y eb , environment
- the setting method includes: digital setting Fixed and imaging setting; the digital setting method of the judgment value of the traffic body in front of the vehicle refers to directly giving estimates to the variables representing each judgment value of the traffic body judgment value array in front of the car, and can optimize and adjust according to the use.
- the imaging setting is to manually capture a video image representing the estimated minimum identifiable traffic body or its benchmark through the vehicle light intelligent control system, and call the following steps S2/S3 in the designated area, that is, the designated area of the image traffic body
- W v is set to be about 1.6 meters long, and the upper part of the vertical bar has a direction towards the car 3-1
- W 1 such as the horizontal flat cylinder in the mark 3-6
- the information in front of the specified vehicle includes 1) the specified vehicle front traffic body contour O v extracted from the image traffic body specified area, the car light contour O 1 and its gray pixel (y 1 ) and the car light area contour O r and its gray pixels (y r ), that is, the information of the traffic body in front of the vehicle; 2) directly extract the gray pixels (y e ) from the video image in front of the vehicle, or extract the outline area of the traffic body in front of the vehicle from the video image in front of the vehicle After the subtraction, the gray pixels in the remaining area are extracted as (y e ), or the gray pixels extracted from the specified area of the image environment are used as (y e ), that is, the environmental information in front of the designated vehicle;
- the inverse imaging principle calculates the following detection values: (1.1) the front edge of the traffic body contour O v in front of the imaging vehicle (contour O v lower edge) or the vehicle The world coordinate distance D m from the front edge of the light profile O 1 (lower edge of the profile O 1 ) to the camera, the world coordinate width W m of the imaged traffic body profile O v , (1.2) the imaged vehicle light profile O 1 The world coordinate vertical width/diameter W 1m and the average value Y 1m of its gray pixels (y 1m ), (1.3) the world coordinate longitudinal span R m of the imaged car light area outline O 1r and its gray pixels (y r)
- 2a) directly use the imaging value as the detection value to replace the above step 2), (1a) according to the position in the image of the above-mentioned imaged traffic body contour O v or car light contour O 1 or car light area contour O 1r , the obtained contour O v Lower edge or the lower edge of contour O 1 or the upper edge of contour O 1a to the lower edge of the image D mp imaging value, contour O v lower edge image width W mp imaging value, image width at the widest point of contour O 1 W 1mp
- the judging rules for judging the category of the traffic body are as follows: (1) vehicle and pedestrian judgment: when the width detection value W m of the specified traffic body, (1.1) is not less than the traffic body digital setting judgment value vehicle width W v , the traffic body Otherwise, if (1.2) is smaller than but not less than the corresponding pedestrian width setting judgment value W w , then it is judged as a pedestrian or a bicycle; otherwise, (1.3) or smaller but not less than the vehicle light
- the diameter/width setting judgment value W 1 , and its light intensity detection value is not less than the car light intensity level setting judgment value Y 1 , then it is judged as a car light, and then it is judged as a vehicle or a motorcycle;
- Car light judgment In strong overcast/dark ambient light intensity, when the light intensity detection value Y m of the designated traffic body outline area is not less than the traffic body digital setting judgment value Y h , and the area vertically
- the span detection value R m is not less than the corresponding vehicle high beam light
- the safety light control logic According to the light intensity level of the obtained vehicle front information environmental light intensity detection value Y em , the vehicle pedestrian and vehicle lights detected by the traffic body information, the safety light control logic generates a control light signal; the safety light control logic includes:
- the flicker includes the number of flickers, a single duration, and an interval duration.
- the number of times is set to 3
- the single duration is 120 milliseconds, which is equivalent to 3 frames of images (PAL standard 25 frames/second), far less than the above-mentioned judgment
- the light takes 1000 milliseconds, and the interval is 320 milliseconds.
- the total time for each flashing light is 1 second.
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
An intelligent control system and method for a vehicle lamp. The intelligent control system for a vehicle lamp comprises a vehicle forward analysis subsystem (1-1), a communication subsystem (1-6), a vehicle lamp control device (1-10), and a safety lamp control logic module embedded therein. The intelligent control method for a vehicle lamp comprises: (1) configuring a detection basic parameter (S1); (2) acquiring a vehicle forward video image (S2); (3) extracting specified vehicle forward information (S3); (4) calculating and determining a specified vehicle forward information feature (S4): ambient light intensity, and vehicle/pedestrian and vehicle lamp light; and (5) a safety lamp control logic controlling a vehicle lamp according to the vehicle forward information feature (S5). The intelligent control system and method for a vehicle lamp implement automatic opening and closing of a proper vehicle lamp in time along with a change of the ambient light intensity, and vehicle, pedestrian and light conditions in front of the vehicle, and can guarantee safe driving.
Description
本发明涉及智能汽车领域,具体地说,车灯的智能控制***及方法。The invention relates to the field of intelligent vehicles, in particular, an intelligent control system and method for vehicle lights.
人们参与交通时,时常出现与车灯有关的危险情况,如,天色变暗时忽略及时开启车灯的车辆,致使其不易被及时准确判定,或是来车开启远光灯长时间照射等非法用灯。设计一个避免这些危险的车灯智能控制***对提高行车安全有着现实意义。When people participate in traffic, there are often dangerous situations related to vehicle lights. For example, vehicles that ignore turning on the lights in time when the sky gets dark make it difficult for them to be accurately judged in time, or illegal vehicles such as turning on high beams for a long time. with lights. Designing an intelligent control system for vehicle lights that avoids these dangers has practical significance for improving driving safety.
发明内容Contents of the invention
本发明的目的是实现安全行车使用车灯。The purpose of the invention is to realize the use of vehicle lights for safe driving.
本发明提出实现上述目标的车灯智能控制***及方法。具体如下:The present invention proposes an intelligent control system and method for vehicle lamps to achieve the above objectives. details as follows:
一种车灯智能控制***,其特征在于包括,车前视频分析子***,安全灯控逻辑模块,通信子***,车灯控制装置;所述车前视频分析子***产生指定车前信息,包括指定车前环境信息或与指定车前交通体信息,所述指定车前信息经过安全灯控逻辑模块后,产生控制车灯信号,通过通信子***连接到车灯控制装置,控制车灯;An intelligent control system for car lights, characterized in that it includes a front-of-car video analysis subsystem, a safety light control logic module, a communication subsystem, and a car light control device; the front-of-car video analysis subsystem generates specified front-of-car information, including Designate the environment information in front of the vehicle or the traffic body information in front of the designated vehicle. After the designated front information passes through the safety light control logic module, it generates a signal to control the vehicle light, and connects to the vehicle light control device through the communication subsystem to control the vehicle light;
所述指定车前环境信息包括指定环境光强类别:不需要开启车灯的白天/需要开启车远光灯的极暗/需要开启车灯的强阴;所述指定车前交通体信息包括交通体类别:车辆/行人包括单车/车远光灯。The specified environmental information in front of the vehicle includes the specified environmental light intensity category: daytime when the lights do not need to be turned on/very dark when the high beam lights need to be turned on/strong shade when the lights need to be turned on; the specified traffic body information in front of the vehicle includes traffic Body category: vehicle/pedestrian including bicycle/vehicle high beam.
根据所述车灯智能控制***,其特征在于,所述通信子***指无线或/或与有线通信子***,包括CAN或/或与蓝牙或/或与WIFI或/或与ModBus或/或与ZigBee。According to the vehicle light intelligent control system, it is characterized in that the communication subsystem refers to a wireless or/or wired communication subsystem, including CAN or/or with Bluetooth or/or with WIFI or/or with ModBus or/or with ZigBee.
根据所述车灯智能控制***,其特征在于,所述安全灯控逻辑模块,嵌入在图像处理子***或在通信子***或在车灯控制装置之中,用硬件或/或与软件实现。According to the vehicle light intelligent control system, it is characterized in that the safety light control logic module is embedded in the image processing subsystem or in the communication subsystem or in the vehicle light control device, and is realized by hardware or/or software.
根据所述车灯智能控制***,其特征在于,所述控制车灯信号包括开启或关闭或闪烁如下车灯及其组合:近光灯,远光灯,示宽灯。According to the vehicle light intelligent control system, it is characterized in that the control vehicle light signal includes turning on or off or flashing the following vehicle lights and combinations thereof: low beam, high beam, and width indicator.
根据所述车灯智能控制***,其特征在于,所述车灯控制装置包括原人工控制车灯优先:临时优先和临时切断;临时优先指只要人操作开启车灯就进入原人工控制车灯状态,该操作完成关闭车灯后,自动恢复本控制;临时切断指需要人工操作另外的切断按键和恢复或重启按键,才能切断和恢复本控制;两者用硬件电路或/或与嵌入式软件实现。According to the vehicle light intelligent control system, it is characterized in that the vehicle light control device includes the original manual control vehicle light priority: temporary priority and temporary cut-off; temporary priority means that as long as the human operation turns on the vehicle light, it will enter the original manual control vehicle light state , after the operation is completed and the lights are turned off, the control is automatically restored; temporary cut-off refers to the need to manually operate another cut-off button and restore or restart button to cut off and restore the control; both are implemented with hardware circuits or/or with embedded software .
一种车灯智能控制方法,其特征在于包括,S1,配置检测基础参数;S2,获取车前视频图像;S3,从所述车前视频图像提取指定车前信息,包括环境信息;S4,计算与判断指定车前信息特征:计算所述指定环境信息光强并判断其光强等级类别;S5,安全灯控逻辑根据所述指定环境光强等级类别,产生控制车灯信号;An intelligent control method for vehicle lights, characterized by comprising: S1, configuring basic parameters for detection; S2, acquiring video images in front of the vehicle; S3, extracting specified information in front of the vehicle from the video images in front of the vehicle, including environmental information; S4, calculating and judging the characteristics of the designated front information of the vehicle: calculating the light intensity of the designated environment information and judging its light intensity level category; S5, the safety light control logic generates a control vehicle light signal according to the specified environment light intensity level category;
所述配置检测基础参数包括标定提供所述车前视频的摄像机内外参数,设定环境光强等级类别的判断值,用以判断环境光强类别是白天/极暗/强阴。The configuration detection basic parameters include calibrating the internal and external parameters of the camera that provides the video in front of the car, and setting the judgment value of the ambient light intensity level category to determine whether the ambient light intensity category is daytime/very dark/strong shade.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S1-1,所述环境光强等级类别设定分为:1.白天,不需要开启车灯的光强;2.极暗,包括夜间无路灯或弱路灯需要开启远光车灯的光强;3.强阴,光强于极暗,包括黄昏/清晨/夜间强路灯/极低云层阴天/隧道内需要开启车灯的光强;各环境光强等级类别或用设定的各对应灰度值作为其判断值;或还设定指定时间间隔的该类别之间灰度值之差最大值,作为判断是否属于环境光强变化差的判断值,小于该值的判定环境光强差,称为环境之差判断值。S1-1, the setting of the ambient light intensity level category is divided into: 1. During the day, the light intensity that does not need to turn on the headlights; 2. Extremely dark, including the light intensity that needs to be turned on when there are no street lights or weak street lights at night. ;3. Strong shade, the light intensity is stronger than extremely dark, including dusk/early morning/night strong street lights/very low clouds overcast/light intensity that needs to be turned on in tunnels; The gray value is used as its judgment value; or the maximum value of the difference between the gray values of the category at the specified time interval is also set as the judgment value for judging whether it belongs to the difference in ambient light intensity change. , called the environment difference judgment value.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S3-1,所述提取指定环境信息的方法指任何提取图像目标背景的算法,包括直接提取图像中所有像素的灰度值,或提取图像中环境指定区域所有像素的灰度值。S3-1, the method for extracting the specified environment information refers to any algorithm for extracting the target background of the image, including directly extracting the gray value of all pixels in the image, or extracting the gray value of all pixels in the specified environment area in the image.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S4-1,所述计算指定环境信息光强包括计算与所提取的指定环境信息即所述灰度值相关的数值,包括其平均值,作为该环境光强特征的检测值。S4-1. The calculating the light intensity of the specified environment information includes calculating a value related to the extracted specified environment information, that is, the gray value, including its average value, as a detection value of the feature of the environment light intensity.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S4-2,所述判断指定环境光强等级类别是将所述指定环境光强检测值与所述设定环境光强等级类别判断值对比,其规则包括:当指定环境光强检测值,(1)不小于设定环境光强等级类别白天判断值时,该环境光强判定为白天;(2)不大于相应极暗判断值时,该环境光强判定为极暗;(3)介于极暗判断值与白天判断值之间时,该环境光强判定为强阴。S4-2, the judging the specified ambient light intensity level category is to compare the specified ambient light intensity detection value with the set ambient light intensity level category determination value, and the rules include: when specifying the ambient light intensity detection value, ( 1) When it is not less than the daytime judgment value of the set ambient light intensity level category, the ambient light intensity is judged as daytime; (2) When it is not greater than the corresponding extremely dark judgment value, the ambient light intensity is judged as extremely dark; (3) between When the extremely dark judgment value is between the daytime judgment value, the ambient light intensity is judged as strong shade.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S4-3,所述判断指定环境光强等级类别是将所述指定环境光强检测值与所述设定环境光强等级类别判断值对比时,参考其之前图像相关光强判定及其与该之前图像相关光强灰度值之差,其规则包括在判断上述规则(1)之前先判断:当指定环境光强检测值,(0)大于其之前图像相关光强类别,其与该之前图像相关光强灰度值之差还不小于环境之差判断值,该环境光强被判定为临时光强;否则,执行上述规则(1)。S4-3. The judging the specified ambient light intensity level category is to compare the specified ambient light intensity detection value with the set ambient light intensity level category judgment value, and refer to the previous image-related light intensity determination and its correlation with the specified ambient light intensity level category. The difference between the previous image-related light intensity gray value, its rule includes judging before the above-mentioned rule (1): when the specified ambient light intensity detection value, (0) is greater than its previous image-related light intensity category, it is different from the previous image If the difference between the relevant light intensity gray values is not less than the environment difference judgment value, the environment light intensity is judged as temporary light intensity; otherwise, the above rule (1) is executed.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S5-1,所述安全灯控逻辑包括:S5-1, the safety light control logic includes:
1)当环境光强为白天,关闭远光灯,近光灯,示宽灯;1) When the ambient light intensity is daytime, turn off the high beam, low beam and width indicators;
2)当环境光强为强阴时,开启/保持近光灯和示宽灯;2) When the ambient light intensity is strong and cloudy, turn on/keep the dipped beam and width indicators;
3)当环境光强为极暗时,开启/保持远光灯。3) When the ambient light intensity is extremely dark, turn on/maintain the high beam.
根据所述车灯智能控制方法,其特征在于,S1a,所述S1中配置检测基础参数还包括设定交通体类别判断值,用以判断交通体类别是车辆/行人/远光灯;S3a,所述S3中从所述车前视频提取的指定车前信息还包括交通体信息;S4a,所述S4中计算与判断指定车前信息特征还包括计算指定交通体信息特征值并判断其类别;S5a,所述S5中安全灯控逻辑还包括根据所述指定车前交通体类别,产生控制车灯信号。According to the intelligent control method for vehicle lights, it is characterized in that, S1a, configuring the detection basic parameters in S1 also includes setting a traffic body type judgment value, to judge whether the traffic body type is vehicle/pedestrian/high beam; S3a, The specified front-of-vehicle information extracted from the front-of-vehicle video in S3 also includes traffic body information; S4a, calculating and judging the features of the specified front-of-vehicle information in S4 also includes calculating the feature value of the specified traffic body information and judging its category; S5a, the safety light control logic in S5 further includes generating a signal to control the vehicle light according to the category of the designated traffic body in front of the vehicle.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S1a-1,所述设定交通体类别判断值是一组估计最小可辨识交通体类别的特征值,包括特征宽度/直径判断值,特征光强等级及其光区纵向跨度判断值;1.车辆宽度:设定以常规最小车辆宽度,并设定其数倍于行人宽度;2.行人宽度:设定以常规最小行人宽度,并设定电单车/摩托车与其同宽度;3.车灯直径:设定以常规最小车灯直径,还设定其与电单车或摩托车灯相同,其光强设定以车灯最小光强;4.车远光灯光强等级:设定以强于近光灯光强,其光区最小纵向/横向跨度以近光灯光区常规最大纵向/横向跨度设定,或者,以设定区域作为远光灯跨度判断设定;5.车近光灯光强等级:设定以强于车灯光强,其光区纵向/横向跨度设定常规最大和最小值;这些判断值可以根据使用情况调整再优化。S1a-1, the set traffic body category judgment value is a set of characteristic values for estimating the smallest identifiable traffic body category, including characteristic width/diameter judgment value, characteristic light intensity level and judgment value of longitudinal span of light zone; 1. Vehicle width: set the conventional minimum vehicle width, and set it several times the pedestrian width; 2. Pedestrian width: set the conventional minimum pedestrian width, and set the motorcycle/motorcycle with the same width; 3. Car lights Diameter: Set the diameter of the conventional minimum car light, and also set it to be the same as the motorcycle or motorcycle light, and set the light intensity to the minimum light intensity of the car light; Low beam light intensity, the minimum vertical/horizontal span of the light area is set by the conventional maximum vertical/horizontal span of the low beam light area, or the set area is used as the judgment setting of the high beam span; 5. Vehicle low beam light intensity level: Set to be stronger than the light intensity of the car, and the vertical/horizontal span of the light zone is set to the conventional maximum and minimum values; these judgment values can be adjusted and optimized according to the usage.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S3a-1,从所述车前视频提取指定车前交通体信息方法包括用图像目标检测算法直接从图像中或从其中交通体指定区域,提取交通体轮廓,获取该轮廓的高、宽或直径尺寸和代表其光强的灰度值;所述图像目标检测算法包括R-CNN算法系列或YOLO算法系列,但不限于这些。S3a-1, the method of extracting the specified traffic body information in front of the vehicle from the video in front of the vehicle includes using the image target detection algorithm to directly extract the contour of the traffic body from the image or from the specified area of the traffic body, and obtain the height, width or diameter of the contour Size and gray value representing its light intensity; the image target detection algorithm includes R-CNN algorithm series or YOLO algorithm series, but not limited to these.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S4a-1,所述计算指定交通体信息特征值包括,提取所述指定交通体轮廓宽度或直径作为其检测值;计算所述指定交通体轮廓光强灰度值的平均值及其纵向/横向跨度作为其检测值。S4a-1, the calculation of the specified traffic body information characteristic value includes extracting the specified traffic body contour width or diameter as its detection value; calculating the average value of the specified traffic body contour light intensity gray value and its longitudinal/transverse direction span as its detection value.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S4a-2,所述判断指定交通体类别包括判断车辆/行人和判断车灯光;将所述交通体轮廓宽度或直径检测值,与所述设定交通体宽度判断值对比,符合交通体判定规则的图像目标,判定为相应交通体:车辆或车灯(也是车辆)或行人或单车;将所述交通体轮廓光强检测值及其纵向/横向跨度检测值,与所述交通体光强等级设定判断值和纵向/横向跨度设定判断值综合对比,符合交通体判定规则的图像目标,判定为相应车辆灯光:车远光灯或车近光灯/车灯。S4a-2, the determination of the specified traffic body category includes judging vehicles/pedestrians and judging vehicle lights; comparing the detection value of the traffic body outline width or diameter with the set traffic body width judgment value, and conforms to the traffic body judgment rules The image target of the image is determined as the corresponding traffic body: vehicle or car light (also a vehicle) or pedestrian or bicycle; The set judgment value is comprehensively compared with the set judgment value of the longitudinal/horizontal span, and the image object conforming to the traffic body judgment rules is judged as the corresponding vehicle light: car high beam or car low beam/light.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S4a-3,所述判断指定交通体类别的交通体判定规则如下:(1)车辆与行人判定:当指定交通体轮廓宽度检测值,(1.1)不小于交通体车辆宽度设定判断值时,该交通体判定为车辆;否则,(1.2)较小的该宽度检测值但不小于相关行人宽度设定判断值时,则判定为行人或单车类;否则,(1.3)或更小的该宽度检测值但不小于相关车灯直径/宽度设定判断值,并且在强阴/极暗环境光强下其光强检测值不小于相应的车灯光强设定判断值,则判定为车灯,进而判为车辆或电单车;(2)车灯光判定:在强阴/极暗环境光强下,当指定交通体轮廓区域光强检测值,(2.1)不小于交通体车远光灯光强设定判断值,同时其区域纵向/横向跨度检测值不小于相关车远光灯光区纵向/横向跨度设定判断值,或者,其区域跨度包括了远光灯跨度判断设定区,则判定为车远光灯;否则,(2.2)较小的该轮廓区光强检测值但不小于相关车近光灯光强设定判断值,或其区域纵向/横向跨度检测值不小于相关车近光灯光区域纵向/横向跨度设定判断值,则判定为车近光灯;S4a-3, the traffic body judging rules for the specified traffic body category are as follows: (1) vehicle and pedestrian judgment: when the detection value of the specified traffic body outline width (1.1) is not less than the traffic body vehicle width setting judgment value, The traffic body is judged to be a vehicle; otherwise, when (1.2) the smaller width detection value is not less than the relevant pedestrian width setting judgment value, it is judged as a pedestrian or bicycle; otherwise, (1.3) or smaller width If the detection value is not less than the judgment value of the diameter/width setting of the relevant car light, and the light intensity detection value is not less than the corresponding car light intensity setting judgment value under strong cloudy/very dark ambient light intensity, then it is judged as a car light , and then judged as a vehicle or a motorcycle; (2) Judgment of vehicle lights: under strong cloudy/very dark ambient light intensity, when the light intensity detection value of the designated traffic body outline area, (2.1) is not less than the traffic body vehicle high beam light intensity Set the judgment value, and at the same time, the detection value of the longitudinal/horizontal span of the area is not less than the set judgment value of the vertical/horizontal span of the high beam light area of the relevant vehicle, or the area span includes the high beam span judgment setting area, then it is judged as car high beam; otherwise, (2.2) the smaller detection value of the light intensity in the outline area is not less than the set judgment value of the low beam light intensity of the relevant vehicle, or the detection value of the longitudinal/transverse span of the area is not less than the low beam light intensity of the relevant vehicle If the judging value is set for the vertical/horizontal span of the area, it is judged to be the low beam of the car;
或上述步骤(2.1)车远光灯的判定还包括如下步骤(2.1a):将(2.1)对车远光灯的判定作为临时判断,将持续了若干连续帧图像的所述临时判断作为远光灯判定,该连续帧图像数所用时长称为判定远光灯需用时长,设定大于估计常规开启远光灯最小时长。Or the judgment of the above-mentioned steps (2.1) the high beam of the car also includes the following steps (2.1a): (2.1) is used as a temporary judgment for the judgment of the high beam of the car, and the temporary judgment that has continued several continuous frame images is taken as a far For light judgment, the time required for the number of consecutive frames of images is called the time required for judging the high beam, and the setting is greater than the estimated minimum time for turning on the high beam.
根据所述车灯智能控制方法,其特征在于,According to the intelligent control method for vehicle lights, it is characterized in that,
S5a-1,所述安全灯控逻辑还包括:S5a-1, the safety light control logic also includes:
4)当环境光强是强阴或极暗,有车辆或行人时,关闭远光灯,保持/开启近光灯;4) When the ambient light intensity is overcast or extremely dark, and there are vehicles or pedestrians, turn off the high beam and keep/turn on the low beam;
5)当环境光强是强阴或极暗,有车辆远光灯时,或启动闪烁远光灯;5) When the ambient light intensity is strong or extremely dark, and there is a vehicle high beam, or start to flash the high beam;
上述闪烁远光灯包括闪烁次数,间隔和单次时长,其单次时长小于判定远光灯需用时长。The flashing high beam above includes the number of flashes, the interval and the duration of a single flash, and the single duration is shorter than the time required for judging the high beam.
本发明优点如下:本发明实现了跟随环境光强变化和车前车辆灯光与行人情况及时自动打开或关闭适当的车灯,即避免暗光环境忘记开启车灯又避免夜间行车远光灯光照射前方车辆行人等情况的事故隐患,保证安全行车节能;人认为必须用或不用车灯时候,可以自主切换;同时简便,车前视频分析子***可嵌入在车辆常用的行车记录仪中。The advantages of the present invention are as follows: the present invention can automatically turn on or off the appropriate car lights in time following the changes in the light intensity of the environment and the vehicle lights and pedestrians in front of the car, that is, to avoid forgetting to turn on the car lights in dark light environments and avoid driving at night. Accident hazards in vehicles and pedestrians, etc., to ensure safe driving and energy saving; when people think it is necessary to use or not use lights, they can switch independently; at the same time, it is simple, and the video analysis subsystem in front of the car can be embedded in the commonly used driving recorder of the vehicle.
图1是车灯智能控制***示意图;Figure 1 is a schematic diagram of an intelligent control system for vehicle lights;
图2是车灯智能控制方法流程图;Fig. 2 is a flow chart of an intelligent control method for vehicle lights;
图3是摄像镜头安装位置示意图;Fig. 3 is a schematic diagram of the installation position of the camera lens;
附图中的编号索引:Numbered index in the attached drawings:
图1:1-1-车前视频分析子***;点划线表示子***,包括1-2至1-5,和1-7;1-2-视频传感装置;1-3-车前视频数据线;1-4-图像处理子***,或包括安全灯控逻辑如标注“灯控逻辑”矩形虚线框;1-5-车前信息或车灯控制数据线;1-6-通信子***,包括1-7至1-9;1-7-无线蓝牙或有线CAN通信装置A端;1-8-通信媒介,无线电波或与有线;1-9-无线蓝牙或有线CAN通信装置B端,包括安全灯控逻辑如标注“灯控逻辑”矩形线框;1-10-车灯控制装置,包括1-9,和1-11至1-17;1-11-车灯控制器,或包括安全灯控逻辑如标注“灯控逻辑”矩形虚线框;1-12-开关控制信号线;1-13-远光灯开关;1-14-远光灯闪烁开关;1-15-近光灯开关;1-16-示宽灯开关;1-17-开关状态传感器控制模块组;1-18-环境信息数据通信媒介,虚线表示或者与视频传感器并用;1-19-环境信息传感器组,虚线表示或者与视频传感器并用;“+”-车灯电源;“S#”-人控车灯开关-常断状态,S13至S16;“S
of”-临时切断开关-常通状态;“S
tof”-临时优先开关-常通状态;“S
cc”-临时切断开关的切断控制操作/信号;图3:3-1-小轿车;3-2-车前风挡窗内后视镜兼行车记录仪显示屏;3-3-车前风挡窗内后视镜背面安装的前视摄像头,高于距地面H
c米;3-4-与摄像机镜头轴线3-5在同一垂面中的水平线,其所在平面中的侧向拉出线3-4″与小轿车纵向对称面和地面的交线3-10在同一平面的侧向拉出线3-10”标注摄像机镜头安装高度H
c;3-5-在小轿车纵向对称面上的摄像机镜头轴对称线,点划线,与水平线3-4形成下俯角γ
c;3-6-由摄像机引出到车道中心线的两 条细虚线表示摄像头纵向视角宽的上下视角线,以及展示横向视角范围宽相关虚线椭圆;3-7-车道线,双线;3-8-交通实体标定T型标杆,其横杆横卧于D
t米距离处地面,呈现倒T形式,横杆长W
v米,立杆长H
1米,立杆上部有车大灯(较大)和示宽灯(较小),示宽灯直径W
1米;3-9和3-10-点划线代表图形中心对称线,包括图3-a中小轿车摄像机显示屏纵向中心对称线和横向中心对称线及它们各自所在平面与本图3中的地面形成相应的交线,两组纵横对称线在图3和图3-a中交叉出两个对称中心点O;3-11-距离车前摄像头D
t米线,粗虚线;3-12-该横线表示地面与天空的分界线,其下面包括道路的区域是图像中交通体指定区域;3-13-图像中环境指定区线,其上至图像上边线的H
e=H
i/8的区域,为图像环境指定区,H
i图像高度;3-14-该粗实线梯形表示该小轿车前舱;其中图3-a--展示小轿车内显示屏上的车前信息;
Figure 1: 1-1-Video analysis subsystem in front of the car; dotted line indicates the subsystem, including 1-2 to 1-5, and 1-7; 1-2-Video sensor device; 1-3-Front of the car Video data line; 1-4-image processing subsystem, or including safety light control logic such as the rectangular dotted box labeled "light control logic"; 1-5-front information or light control data line; 1-6-communicator System, including 1-7 to 1-9; 1-7-wireless Bluetooth or wired CAN communication device A; 1-8-communication medium, radio wave or wired; 1-9-wireless Bluetooth or wired CAN communication device B terminal, including safety light control logic, such as the rectangular wire frame marked "light control logic"; 1-10-car light control device, including 1-9, and 1-11 to 1-17; 1-11-car light controller, Or include safety light control logic such as marked "light control logic" rectangular dotted line box; 1-12-switch control signal line; 1-13-high beam switch; 1-14-high beam flashing switch; 1-15-near Light switch; 1-16-width light switch; 1-17-switch state sensor control module group; 1-18-environmental information data communication medium, dotted line indicates or used together with video sensor; 1-19-environmental information sensor group , the dotted line indicates or is used together with the video sensor; "+"-light power supply; "S#"-human control light switch-normal off state, S13 to S16; "S of "-temporary cut-off switch-normal on state;" S tof ”-temporary priority switch-normally on state; “S cc ”-cut-off control operation/signal of temporary cut-off switch; Figure 3: 3-1-car; 3-2-windshield and interior rearview mirror The display screen of the driving recorder; 3-3-the front-view camera installed on the back of the rear-view mirror in the front windshield window of the car, which is higher than H c meters from the ground; 3-4-and the camera lens axis 3-5 in the same vertical plane Horizontal line, the intersection line 3-10 of the lateral pull-out line 3-4 "in the plane where it is located and the longitudinal symmetry plane of the car and the ground" marks the installation height Hc of the camera lens on the side pull-out line 3-10" of the same plane; 3 -5- The axisymmetric line of the camera lens on the longitudinal symmetry plane of the car, the dot-dash line, and the horizontal line 3-4 form a depression angle γc ; 3-6- The two thin dotted lines drawn from the camera to the center line of the lane represent the camera Longitudinal viewing angle wide up and down viewing angle lines, and related dotted line ellipse showing wide horizontal viewing angle range; 3-7-lane line, double line; 3-8-traffic entity calibration T-shaped pole, whose crossbar lies at a distance of D t meters The ground is in the form of an inverted T, the length of the horizontal bar is W v meters, the length of the vertical pole is 1 meter, and there are headlights (larger) and width lights (smaller) on the upper part of the pole, and the diameter of the width light is W 1 meter; 3 -9 and 3-10-The dotted line represents the center line of symmetry of the figure, including the longitudinal center line of symmetry and the transverse center line of symmetry of the car camera display in Figure 3-a and their respective planes corresponding to the ground in Figure 3 Intersection line, two sets of vertical and horizontal symmetry lines intersect two symmetrical center points O in Fig. 3 and Fig. 3-a; 3-11-distance from the front camera D t meter line, thick dotted line; 3-12-The horizontal line represents the ground The boundary line with the sky, the area including the road below it is the designated area of the traffic body in the image; 3-13-the designated area line of the environment in the image, the area of He = H i /8 up to the upper edge of the image is the image Designated environment area, H i image height; 3-14-the thick solid line trapezoid represents the front cabin of the car; Figure 3-a-shows the front information on the display screen in the car;
结合附图详细描述本发明车灯智能控制***一种实施例:An embodiment of the vehicle light intelligent control system of the present invention is described in detail in conjunction with the accompanying drawings:
如图1,取一个配置了充足内存、运算处理车前视频图像能力的行车视频记录仪开发板作为车前视频分析子***如图1标注1-1,该板上的摄像镜头模块作为如标注1-2的所述视频传感器,该模块车前视频数据线作为如1-3数据线,该板上的以处理器模块组为中心的数据处理子***作为图像处理子***如1-4,1-2所获取车前视频图像通过车前视频数据线1-3经过图像处理子***处理1-4处理后,提取出包括指定车前环境信息特征和指定车前交通体信息类别,通过车前信息数据线1-5和通信子***1-6,其B端嵌入的安全灯控逻辑产生车灯控制信号,送给车灯控制装置1-10,其中车灯控制器1-11将车灯控制信号1-12送给车灯开关1-13至1-16;As shown in Figure 1, take a driving video recorder development board equipped with sufficient memory and the ability to calculate and process video images in front of the car as the video analysis subsystem in front of the car. The video sensor of 1-2, the front video data line of the module is used as the data line 1-3, and the data processing subsystem centered on the processor module group on the board is used as the image processing subsystem such as 1-4, 1-2 The acquired front-of-car video image is processed by the image processing subsystem 1-4 through the front-of-car video data line 1-3, and the features of the environment information in front of the car and the traffic body information category in front of the car are extracted. The front information data line 1-5 and the communication subsystem 1-6, the safety light control logic embedded in the B terminal generates the car light control signal, which is sent to the car light control device 1-10, wherein the car light controller 1-11 controls the car light Light control signal 1-12 is sent to car light switch 1-13 to 1-16;
所述指定车前环境信息特征包括:(e1)白天,不需要开启车灯;(e2)极暗,包括夜间无路灯或弱路灯需要开启远光车灯;(e3)强阴,包括黄昏,清晨,夜间强路灯,极低云层阴天,或隧道内需要开启车灯;The characteristics of the specified front environment information include: (e1) During the day, it is not necessary to turn on the lights; (e2) It is extremely dark, including no street lights or weak street lights at night, and the high beam lights need to be turned on; (e3) Strong shade, including dusk, In the early morning, strong street lights at night, extremely low clouds and cloudy days, or the lights need to be turned on in the tunnel;
所述指定车前交通体信息类别包括(v1)车辆和行人或单车;(v2)车灯近光光强介于强阴与白天之间;(v3)车灯远光灯光强强于白天;The specified traffic body information category in front of the vehicle includes (v1) vehicles and pedestrians or bicycles; (v2) the light intensity of the low beam of the vehicle lights is between strong shade and daytime; (v3) the intensity of the high beam lights of the vehicle lights is stronger than that of the daytime;
所述通信子***1-6包括内置于所述开发板***内的蓝牙通信模块作为该通信子***A端1-7,和用传输媒介1-8无线微波2.4GHz连接的置于车灯控制装置1-10中的蓝牙通信模块nRF52832作为该通信子***B端1-9,传输给1-10;The communication subsystem 1-6 includes a bluetooth communication module built in the development board system as the communication subsystem A terminal 1-7, and a wireless microwave 2.4GHz connection with a transmission medium 1-8 placed in the car light control The Bluetooth communication module nRF52832 in the device 1-10 is used as the B terminal 1-9 of the communication subsystem, and is transmitted to 1-10;
所述车灯控制装置1-10,从其蓝牙通信模块B端1-9接收到所述A端1-7送过来的车前信息数据1-5,通过安全灯控逻辑把车前信息数据变成为车灯控制信号,经过车灯控制器1-11送给车灯开关:(1)远光灯如1-13,(2)远光灯闪烁(3次,间隔320毫秒,单次时长120毫秒)如1-14,(3)近光灯如1-15,(4)示宽灯如1-16;所述车灯控制器1-11也可以嵌入安全灯控逻辑(如标注“灯控逻辑”的虚线框)取代嵌入蓝牙通信模块B端1-9;另外,或与参考开关电流传感器控制模块组如图标1-17获取的上述各车灯人工操作开关状态信息:(1s.1)远光灯 开或者关,(1s.2)远光灯闪烁开或者关,(1s.3)近光灯开或者关,(1s.4)示宽光灯开或者关,同时,为1-11的电源串接装两个常通状态开关,临时切断开关S
of和临时优先开关S
tof,S
cc是S
tof切断控制操作/信号;
The vehicle light control device 1-10 receives the front information data 1-5 sent by the A terminal 1-7 from its Bluetooth communication module B terminal 1-9, and transmits the front information data 1-5 through the safety light control logic. Become a car light control signal, and send it to the car light switch through the car light controller 1-11: (1) the high beam light is like 1-13, (2) the high beam light flashes (3 times, the interval is 320 milliseconds, a single duration 120 milliseconds) such as 1-14, (3) dipped beam such as 1-15, (4) width indicator light such as 1-16; the car light controller 1-11 can also be embedded in safety light control logic (as marked The dotted line box of "light control logic" replaces the B-terminal 1-9 embedded in the Bluetooth communication module; in addition, or with the reference switch current sensor control module group as shown in the icon 1-17 to obtain the above-mentioned manual operation switch status information of each car light: (1s .1) High beam is on or off, (1s.2) high beam is flashing on or off, (1s.3) low beam is on or off, (1s.4) wide beam is on or off, and at the same time, Install two normally-on switches in series for the power supply of 1-11, a temporary cut-off switch S of and a temporary priority switch S tof , and S cc is the S tof cut-off control operation/signal;
所述车灯控制器1-11包括器件Darlington-ULN2003和固态继电器SSR模块将3v电信号控制12v至24v电路电器,或还包括替代B端嵌入的安全灯控逻辑;The car light controller 1-11 includes a device Darlington-ULN2003 and a solid-state relay SSR module to control 12v to 24v circuit electrical appliances with a 3v electrical signal, or also includes a safety light control logic embedded in the B terminal;
所述车前视频分析子***或与配置光传感器模块组如图标1-18,1-19,作为另外一个环境信息获取配置选项,以获取环境信息特征(e1)环境光强是白天;(e2)环境光强是极暗,如弱路灯或无路灯;(e3)环境光强是强阴,如黄昏或清晨或夜间强路灯;(e4)环境光强是(v2)近光灯;(e5)环境光强是(v3)远光灯;送往开发板***或直接送到通信子***B端;The front-of-the-car video analysis subsystem or the light sensor module group configured as icons 1-18, 1-19, as another environmental information acquisition configuration option, to obtain environmental information characteristics (e1) the ambient light intensity is daytime; (e2 ) the ambient light intensity is extremely dark, such as weak street lights or no street lights; (e3) the ambient light intensity is strong shade, such as dusk or early morning or strong street lights at night; (e4) the ambient light intensity is (v2) dipped beam; (e5) ) Ambient light intensity is (v3) high beam; sent to the development board system or directly sent to the B terminal of the communication subsystem;
所述车灯控制器包括原人工车灯开关电路优先的临时切断和临时优先电路:将本车灯控制器与原人工车灯开关电路并联,并在本控制器中串接2个“常通状态”开关;其中一个,S
of,用于切断本控制器,用作临时切断;另一个,S
tof,受控于1-17的接通信号S
cc,与原人工车灯开关作反操作联动,即当原人工接通开启车灯时,切断本控制器,当原人工切断关闭车灯时,恢复接通本控制器,即临时优先。
The car light controller includes a temporary cut-off and a temporary priority circuit for the priority of the original manual car light switch circuit: the car light controller is connected in parallel with the original manual car light switch circuit, and two "normally on" circuits are connected in series in this controller. State"switch; one of them, S of , is used to cut off the controller, and is used as a temporary cut-off; the other, S tof , is controlled by the connection signal S cc of 1-17, which is reversed to the original manual light switch Linkage, that is, when the lights are turned on manually, the controller is cut off, and when the lights are turned off manually, the controller is turned on again, that is, temporary priority.
本发明车灯智能控制方法一种实施例,其方法流程如图2,An embodiment of the vehicle lamp intelligent control method of the present invention, the method flow chart is shown in Figure 2,
S1e,配置检测基准参数:标定所述车灯智能控制***的视频传感器-摄像镜头内外参数,设定车前信息判断值,包括设定车前环境光强类别等级判断值和设定车前交通体类别判断值;S1e, configure the detection reference parameters: calibrate the internal and external parameters of the video sensor-camera lens of the intelligent control system for vehicle lights, set the judgment value of the information in front of the vehicle, including setting the judgment value of the environmental light intensity in front of the vehicle and the traffic in front of the vehicle Judgment value of body category;
所述标定摄像机内外参数;外参数包括(1)在小轿车如图3标注3-1内前风挡玻璃内的后视镜如标注3-2背面安装摄像机,其镜头3-3指向车前,其水平线3-4所在平面中的侧向拉出线3-4″与小轿车纵向对称面和地面的交线3-10在同一平面的侧向拉出线3-10″标注摄像机镜头安装距离地面高度H
c,约1.5米;(2)调整固定该摄像机镜头使得摄像画面中地面与天空的分界线位置3-12,本实施例使得距离摄像机镜头的100米线D
t与该画面水平中线3-9重合,获取摄像机镜头中心轴线3-5与其水平线3-4的下俯角γ
c,得到上述分界线位于靠近摄像画面水平中线偏上处如标注3-12,其下面至图像下边缘包括道路的区域是图像中交通体指定区;其上包括图像环境指定区,本实施例以线3-13标注该线上至图像上边线的1/8图像区域;选用***,其内参数包括镜头主距f等6个,可安装前标定好;所述内外参数均可在本产品出厂时预先计算设定好,以简单选择方式提供车辆用户或4S店配置设定;
The internal and external parameters of the calibration camera; the external parameters include (1) the rearview mirror in the front windshield of the car as shown in Figure 3 marked 3-1 as marked 3-2 back to install the camera, its lens 3-3 points to the front of the car, The lateral pull-out line 3-4″ in the plane where the horizontal line 3-4 is located and the intersection line 3-10 of the car’s longitudinal symmetry plane and the ground are on the same plane. H c , about 1.5 meters; (2) adjust and fix the camera lens so that the boundary position between the ground and the sky in the camera picture is 3-12, and this embodiment makes the 100-meter line D t from the camera lens and the horizontal center line of the picture 3-9 Coincidentally, obtain the depression angle γ c between the central axis 3-5 of the camera lens and the horizontal line 3-4, and obtain the above-mentioned dividing line located close to the upper part of the horizontal center line of the camera picture, such as marked 3-12, and the area from below to the lower edge of the image including the road It is the designated area of the traffic body in the image; it includes the designated area of the image environment, and the present embodiment marks the 1/8 image area from this line to the upper edge of the image with line 3-13; select a pinhole camera for use, and its internal parameters include the main lens There are 6 distances such as f, which can be calibrated before installation; the internal and external parameters can be pre-calculated and set when the product leaves the factory, and provide vehicle users or 4S shop configuration settings in a simple selection method;
所述车前环境光强类别包括白天,极暗,强阴,其光强等级设定判断值包括白天Y
ed,极暗Y
eb,环境之差ΔY
ex(ΔT);所述环境之差ΔY
ex(ΔT)是指定时间间隔ΔT内的指定环境光强类别之间灰度值之差最大判断值,小于该差值的认为是环境光强差,大于该差值的增强的是临时光强差;用摄像视频图像的设定灰度值代表,其设定方式包括数字设定,成像设定和自动设定;所述数字设定指直接给代表白天Y
ed,极暗Y
eb,环境之差判断值ΔY
ex(ΔT)的变量赋予0至255之间的整数值,如Y
ed=130,Y
eb=70,ΔY
ex(ΔT)=5,ΔY
ex(ΔT)的指定时间间隔是帧图像时长40毫秒(PAL标准25帧/秒),并可根据使用经验或智能算法调整;所述成像设定是通 过所述车灯智能控制***的人为摄取实景环境画面,调用下面步骤S2/S3中相关程序计算出相关灰度值,赋予上述变量作为设定判断值,包括:(1)手动摄取人认为不需要开启近光灯的最小环境光强的视频图像,所述***自动计算其指定区域平均灰度值Y
ed,保存为所述环境光强等级白天的设定判断值Y
ed;(2)将同(1)理获取的需要开启远光灯的环境视频图像指定区域平均灰度值Y
eb,保存为所述环境光强等级极暗的设定判断值Y
eb;(3)将同(1)理获取从早晨强阴随时间t直到需要开启车灯时的白天期间ΔT的逐渐增强的k帧环境视频图像指定区域平均灰度值
计算它们相邻帧
的差绝对值
找出其中最大差
存为所述环境之差判断值ΔY
ex(ΔT);以上步骤,可以在本产品出厂时操作或在4S店操作;所述自动设定是自动成像设定,在实际驾驶车辆中分别自动保存人工首次或多次进行如上述成像设定中的(1)或(2)或(3)操作时的视频图像设定区域平均灰度值,或多次该区域平均灰度值的相关平均灰度值,作为上述相应的环境光强等级设定判断值;所述环境视频图像指定区域,简称为图像环境指定区,是为减少包括自车灯在内的车灯光对环境光强判定的影响,可设定在所述摄像画面中远离地面一个或若干小区域,本实施例将其设在图像上方1/8区域;
The environmental light intensity category in front of the car includes daytime, extremely dark, and strong shade, and its light intensity level setting judgment value includes daytime Y ed , extremely dark Y eb , and the environmental difference ΔY ex (ΔT); the environmental difference ΔY ex (ΔT) is the maximum judging value of the gray value difference between the specified ambient light intensity categories within the specified time interval ΔT, the value smaller than this difference is regarded as the ambient light intensity difference, and the enhancement greater than the difference value is temporary light intensity It is represented by the set gray value of the camera video image, and its setting methods include digital setting, imaging setting and automatic setting; the digital setting refers to directly representing daytime Y ed , extremely dark Y eb , environment The variable of the difference judgment value ΔY ex (ΔT) is assigned an integer value between 0 and 255, such as Y ed =130, Y eb =70, ΔY ex (ΔT)=5, and the specified time interval of ΔY ex (ΔT) is The frame image duration is 40 milliseconds (PAL standard 25 frames/second), and can be adjusted according to experience or intelligent algorithms; the imaging setting is to artificially capture the real scene environment picture through the intelligent control system of the car lights, and call the following steps S2/ The relevant program in S3 calculates the relevant gray value, and assigns the above-mentioned variable as a set judgment value, including: (1) manually ingesting a video image of the minimum ambient light intensity that people think does not need to turn on the low beam, and the system automatically calculates its The average gray value Y ed of the specified area is saved as the set judgment value Y ed of the daytime environment light intensity level; The degree value Y eb is saved as the set judgment value Y eb of the extremely dark ambient light intensity level; (3) In the same way as (1), obtain the daytime period ΔT from strong shade in the morning over time t to the time when the lights need to be turned on The average gray value of the specified area of the gradually enhanced k-frame environmental video image count their adjacent frames The absolute value of the difference find the biggest difference Stored as the environmental difference judgment value ΔY ex (ΔT); the above steps can be operated when the product leaves the factory or in a 4S shop; the automatic setting is an automatic imaging setting, which is automatically saved in the actual driving vehicle The average gray value of the video image setting area when the operation of (1) or (2) or (3) in the above imaging setting is performed manually for the first time or multiple times, or the related average gray value of the average gray value of the area for multiple times As the above-mentioned corresponding ambient light intensity level setting judgment value; the designated area of the environmental video image, referred to as the image environment designated area for short, is to reduce the influence of the car lights including the own car lights on the judgment of the ambient light intensity , it can be set to one or several small areas away from the ground in the camera picture, which is set in the upper 1/8 area of the image in this embodiment;
所述车前交通体设定判断值是一组对交通体的估计最小可辨识特征值名称及其值J[*],它包括J[小轿车宽度W
v=?,行人宽度W
w=?,{车灯直径/宽度W
1=?,及其光强Y
1=?},{车远光灯光强Y
h=?,及其光区纵向/横向跨度R
h=?},{车近光灯光强Y
11=?,及其光区纵向/横向跨度R
11=?,}],以宽度/直径/跨度/光强,辨识车辆,行人或单车,车灯,远光灯和近光灯,“=?”代表设定判断值;其设定方法包括:数字设定和成像设定;所述车前交通体判断值数字设定方法指直接给代表车前交通体判断值数组各判断值的变量赋予估计各相应数值,并可根据使用优化调整,本实施例对上述设定判断值的数字赋值是:J
ed[小轿车宽度W
v=1.6米,行人宽度W
w=W
v/4,{车灯直径/宽度W
1=Wv/16及其光强Y
1大于车灯最小光强灰度值,160},{车远光灯光强Y
h大于车近光灯光强Y
11,及其光区跨度R
h>=5.0米,或者,其设定区,如将环境指定区设定为远光灯跨度设定区},{车近光灯光强Y
11>Y
1,及其光区跨度R
11>=1.0米}];
The set judgment value of the traffic body in front of the vehicle is a group of estimated minimum identifiable characteristic value names of the traffic body and its value J[*], which includes J[car width W v =? , Pedestrian width W w =? , {car lamp diameter/width W 1 =? , and its light intensity Y 1 =? }, {Car high beam light intensity Y h =? , and the vertical/horizontal span of the light zone R h =? }, {vehicle low beam light intensity Y 11 =? , and its longitudinal/horizontal span of the light zone R 11 =? ,}], with width/diameter/span/light intensity, to identify vehicles, pedestrians or bicycles, lights, high beams and low beams, "=?" represents the set judgment value; the setting method includes: digital setting Fixed and imaging setting; the digital setting method of the judgment value of the traffic body in front of the vehicle refers to directly giving estimates to the variables representing each judgment value of the traffic body judgment value array in front of the car, and can optimize and adjust according to the use. This embodiment The digital assignment of the above-mentioned set judgment value is: J ed [car width W v = 1.6 meters, pedestrian width W w = W v /4, {car light diameter/width W 1 = Wv/16 and its light intensity Y 1 is greater than the minimum light intensity gray value of the car light, 160}, {the light intensity Y h of the high beam of the car is greater than the light intensity Y 11 of the low beam of the car, and the span of the light area R h >= 5.0 meters, or, its setting area, For example, the designated environment area is set as the high beam span setting area}, {vehicle low beam light intensity Y 11 >Y 1 , and its light area span R 11 >=1.0m}];
所述成像设定是通过所述车灯智能控制***,人工手动摄取代表估计的最小可辨识交通体或其标杆的视频图像,在其指定区域即图像交通体指定区,调用下面步骤S2/S3中相关程序提取出相应若干(如20个)距离上最小成像值,每两个D
tp间之间距差成像ΔD
tp=图像高/2/20,形成基于J[*]和距离D
t及成像值D
tp的20个成像设定判断值J
ep[(D
t,D
tp),(J
ed[*],J
ep[*])],其中,J
ep[小轿车宽度W
vp,行人宽度W
wp,{车灯直径/宽度W
1p及其光强Y
1p},{车远光灯光强Y
hp及其光区轮廓纵向跨度R
hp},{车近光灯光强Y
11p及其光区轮廓纵向跨度R
11p}];其步骤包括:(1)在距离摄像机前方设定距离D
t处如图3标注3-11,比如设定D
t约为100米处,路面上倒立放一T-型标杆如标注3-8,横杆在下横卧触地,其横杆代表小轿车轮廓最小宽度W
v,如设定W
v长约1.6米,和立杆上部有一朝向小轿车3-1的0.1米直径W
1的车灯如标注3-6中的横扁圆柱,获取该距离该标杆在摄像机画面中交通体轮廓O
v的宽度W
p成像值W
vp和车灯光轮廓0
1的直径/宽度W
1p;同理,计算获取标杆轮廓0
v前边与摄像头之间距离D
t的成像值D
tp, 再分别打开示宽灯,远光灯和近光灯,计算获取车灯轮廓O
1的灰度值Y
1及其成像值及Y
1p,车远光灯区轮廓O
Rh的最小纵向跨度R
h及其灰度值Y
h的的各自成像值R
hp及Y
hp,车近光灯区轮廓O
R1的最小纵向跨度R
1及其灰度值Y
11的各自成像值R
11p及Y
11p;(2)同理,获取所述预先选择若干(20个)距离位置上的成像判断值,将J
ep[(D
t,D
tp),(J
ed[*],J
ep[*])]中物-像对J
ed[*]和J
ep[*]对应项组合,得到交通体成像设定判断值数组J
ep[(D
t,D
tp),(W
v,W
vp),(W
w=W
v/4,W
wp=W
vp/4),(W
1=W
v/16,W
1p=W
vp/16),(Y
1,Y
1p)},{(Y
h,Y
hp),(R
h,R
hp)},{(Y
11,Y
11p),(R
11,R
11p)}];以上步骤也用为本产品出厂配置或4S店配置设定判断值;所述其指定区域,称为图像交通体指定区,是为减少环境光强对交通体判定的影响,可设定在所述摄像画面中只出现交通体地面相关区域;本实施例的图像交通体指定区设定在图像中天地边际线如图标3-12下方仅含道路区域;
The imaging setting is to manually capture a video image representing the estimated minimum identifiable traffic body or its benchmark through the vehicle light intelligent control system, and call the following steps S2/S3 in the designated area, that is, the designated area of the image traffic body The middle correlation program extracts the minimum imaging values corresponding to several (such as 20) distances, and the distance difference imaging ΔD tp between every two D tp = image height/2/20, forming an image based on J[*] and distance D t and imaging 20 imaging setting judgment values of value D tp J ep [(D t , D tp ), (J ed [*], J ep [*])], where J ep [car width W vp , pedestrian width W wp , {car lamp diameter/width W 1p and its light intensity Y 1p }, {vehicle high beam light intensity Y hp and its longitudinal span R hp of the outline of the light area R hp }, {vehicle low beam light intensity Y 11p and its light area Outline longitudinal span R 11p }]; the steps include: (1) set the distance D t in front of the camera as shown in Figure 3 and mark 3-11, for example, set D t to be about 100 meters, and place a T upside down on the road -Type benchmarks, such as marked 3-8, the horizontal bar is lying on the bottom and touches the ground, and its horizontal bar represents the minimum width W v of the outline of the car. If W v is set to be about 1.6 meters long, and the upper part of the vertical bar has a direction towards the car 3-1 For a car light with a diameter of 0.1 m W 1 such as the horizontal flat cylinder in the mark 3-6, obtain the distance W p imaging value W vp of the traffic body contour O v and the diameter of the car light contour 0 1 in the camera picture /Width W 1p ; similarly, calculate and obtain the imaging value D tp of the distance D t between the front edge of the pole outline 0 v and the camera, and then turn on the width light, high beam and low beam respectively, and calculate and obtain the vehicle light outline O 1 The gray value Y 1 and its imaging value and Y 1p , the minimum longitudinal span R h of the outline O Rh of the high beam area of the vehicle and the respective imaging values R hp and Y hp of the gray value Y h , the low beam of the vehicle The minimum longitudinal span R 1 of the area profile O R1 and the respective imaging values R 11p and Y 11p of the gray value Y 11 ; (2) similarly, obtain the imaging judgment values at several (20) distance positions selected in advance , combine the object-image pairs J ed [*] and corresponding items of J ep [*] in J ep [(D t , D tp ), (J ed [*], J ep [ *])] to obtain the traffic volume Imaging setting judgment value array J ep [(D t , D tp ), (W v , W vp ), (W w =W v /4, W wp =W vp /4), (W 1 =W v / 16, W 1p =W vp /16), (Y 1 , Y 1p )}, {(Y h , Y hp ), (R h , R hp )}, {(Y 11 , Y 11p ), (R 11 , R 11p )}]; the above steps are also used to set the judgment value for the factory configuration of this product or the configuration of the 4S shop; the designated area is called the designated area of the image traffic body, which is used to reduce the ambient light intensity to judge the traffic body Influence, it can be set that only relevant areas on the ground of the traffic body appear in the camera picture; the designated area of the traffic body in the image of the present embodiment is set at the boundary line of the sky and earth in the image, such as below the icon 3-12, which only contains the road area;
注:设立图像交通体指定区,图像指定环境区,和预设若干成像设定判断值,其每一个都提供了降低对图像处理器运算速度能力依赖要求的优势特征,使得本方法可以用较低速度图像处理器,即较低成本的实现;Note: set up image traffic body designated area, image designated environment area, and preset several imaging setting judgment values, each of which provides the advantage of reducing the dependence on the image processor's computing speed capability, so that this method can be used more Low-speed image processors, i.e. lower-cost implementations;
S2e,使用上述摄像镜头模块获取车前视频图像;S2e, using the above-mentioned camera lens module to obtain the video image in front of the car;
S3e,从上述车前视频图像的设定区域提取指定车前信息,使用软件模块或/或与FPGA硬件模块展开的提取图像目标及其轮廓的目标检测算法,如R-CNN算法系列或YOLO算法系列;该指定车前信息包括1)从所述图像交通体指定区提取的指定车前交通体轮廓O
v及车灯轮廓O
1及其灰色像素(y
1)及其车灯光区轮廓O
r及其灰色像素(y
r),即车前交通体信息;2)直接从上述车前视频图像中提取灰色像素(y
e),或将所述车前交通体轮廓区域从上述车前视频图像中减去后提取所余区域的灰色像素作为(y
e),或从所述图像环境指定区提取的灰色像素作为(y
e),即指定车前环境信息;
S3e, extracting the designated front information from the set area of the above-mentioned front video image, using a software module or/or a target detection algorithm for extracting the image target and its outline developed with the FPGA hardware module, such as the R-CNN algorithm series or the YOLO algorithm series; the information in front of the specified vehicle includes 1) the specified vehicle front traffic body contour O v extracted from the image traffic body specified area, the car light contour O 1 and its gray pixel (y 1 ) and the car light area contour O r and its gray pixels (y r ), that is, the information of the traffic body in front of the vehicle; 2) directly extract the gray pixels (y e ) from the video image in front of the vehicle, or extract the outline area of the traffic body in front of the vehicle from the video image in front of the vehicle After the subtraction, the gray pixels in the remaining area are extracted as (y e ), or the gray pixels extracted from the specified area of the image environment are used as (y e ), that is, the environmental information in front of the designated vehicle;
S4e,计算判断指定车前信息中的检测值:1)计算判断上述指定车前环境信息的检测值:(1)计算上述指定车前环境信息(y
e)在指定区域平均灰度值Y
em作为其光强检测值;(2)根据环境光强等级设定判断值和环境之差判断值ΔY
ex(ΔT),对比该光强检测值Y
em进行判定,得出对车前环境光强判断:白天,极暗,强阴,或临时光强;其判定规则如下,当指定环境光强检测值Y
em,(2.0)大于其之前环境光强类别,其与该之前环境光强灰度值之差还不小于环境之差判断值ΔY
ex(ΔT),该环境光强被判定为临时光强;否则,(2.1)不小于环境光强等级白天设定判断值Y
ed时,该环境光强被判定为白天;(2.2)不大于极暗设定判断值Y
eb时,被判定为极暗;(2.3)介于极暗设定判断值Y
eb与白天设定判断值Y
ed之间时,被判定为强阴;
S4e. Calculating and judging the detection value in the information in front of the specified vehicle: 1) Calculating and judging the detection value of the environmental information in front of the designated vehicle: (1) Calculating the average gray value Y em of the specified area in front of the environmental information (y e ) As its light intensity detection value; (2) according to the difference judgment value ΔY ex (ΔT) of setting judgment value and environment according to the environment light intensity level, compare this light intensity detection value Y em to judge, draw the environmental light intensity in front of the car Judgment: daytime, extremely dark, overcast, or temporary light intensity; the judgment rules are as follows, when the specified ambient light intensity detection value Y em , (2.0) is greater than its previous ambient light intensity category, it is the same as the previous ambient light intensity grayscale value difference is not less than the environmental difference judgment value ΔY ex (ΔT), the ambient light intensity is judged as temporary light intensity ; The light intensity is judged as daytime; (2.2) When it is not greater than the extremely dark setting judgment value Y eb , it is judged as extremely dark; (2.3) It is between the extremely dark setting judgment value Y eb and the daytime setting judgment value Y ed time, it is judged as strong Yin;
2)计算判断上述指定车前交通体信息的检测值:(1)根据上述成像的交通体轮廓0v或与车灯轮廓O
1及其车灯光区轮廓O
1r在图像中位置,和摄像机标定内外参数与摄像几何光学成像原理及其世界坐标与成像坐标变换原理,逆成像原理计算下述检测值:(1.1)所述成像车前交通体轮廓O
v前沿(轮廓O
v下边缘)或与车灯轮廓O
1前沿(轮廓O
1下边缘)到所述摄像机的世界坐标距离D
m,所述成像交通体轮廓O
v的世界坐标宽度W
m,(1.2)所述成像车灯光轮廓O
1的世界坐标纵宽度/直径W
1m及其灰色像素(y
1m)的平均值Y
1m,(1.3)所述成像车灯光区轮廓O
1r的世界 坐标纵向跨度R
m及其灰色像素(y
rm)的平均值Y
rm,得到检测值数组M(D
m,W
m,W
1m,Y
1m,R
m,Y
rm);注:成像车灯光区跨度R
m会很大因为远光灯光能充满屏幕;(2)比较上述检测值数组和交通体判断值数字设定数组J
ed[*],判断交通体类别;
2) Calculate and judge the detection value of the traffic body information in front of the designated vehicle: (1) According to the above imaged traffic body contour 0v or the position in the image with the car light contour O 1 and the car light area contour O 1r , and the inside and outside of the camera calibration Parameters and camera geometry optics imaging principle and its world coordinates and imaging coordinate transformation principle, the inverse imaging principle calculates the following detection values: (1.1) the front edge of the traffic body contour O v in front of the imaging vehicle (contour O v lower edge) or the vehicle The world coordinate distance D m from the front edge of the light profile O 1 (lower edge of the profile O 1 ) to the camera, the world coordinate width W m of the imaged traffic body profile O v , (1.2) the imaged vehicle light profile O 1 The world coordinate vertical width/diameter W 1m and the average value Y 1m of its gray pixels (y 1m ), (1.3) the world coordinate longitudinal span R m of the imaged car light area outline O 1r and its gray pixels (y rm ) Average value Y rm , get detection value array M(D m , W m , W 1m , Y 1m , R m , Y rm ); Note: The span R m of the imaging car light area will be large because the high beam light can fill the screen; (2) compare the above-mentioned detection value array and the traffic body judgment value digital setting array J ed [*], judge the traffic body category;
或者,2a)直接用成像值作为检测值替代上述步骤2),(1a)根据上述成像交通体轮廓O
v或车灯轮廓O
1或车灯光区轮廓O
1r在图像中位置,获取的轮廓O
v下边缘或轮廓O
1的下边缘或轮廓O
1a的上边缘到图像下边沿的距离D
mp成像值,轮廓O
v下边缘图像宽度W
mp成像值,轮廓O
1最宽处图像宽度W
1mp成像值及其光强成像值Y
1mp,轮廓O
1r跨度R
mp成像值及其光强Y
rmp成像值,即成像检测值组M
p(D
mp,W
mp,W
1mp,Y
1mp,R
mp,Y
rmp);(2a)将上述成像检测值与该成像距离D
mp=<D
tp但D
mp>D
tp+ΔD
tp的D
tp所对应交通体判断值成像值组J
ep[(D
t,D
tp),(W
v,W
vp),(W
w=W
vp/4,W
wp=W
vp/4),{(W
1=W
vp/16,W
1p=W
vp/16),(Y
1,Y
1p)},{(Y
h,Y
hp),(R
h,R
hp)},{(Y
1,Y
1p),(R
1,R
1p)}]比较,判断交通体类别;
Or, 2a) directly use the imaging value as the detection value to replace the above step 2), (1a) according to the position in the image of the above-mentioned imaged traffic body contour O v or car light contour O 1 or car light area contour O 1r , the obtained contour O v Lower edge or the lower edge of contour O 1 or the upper edge of contour O 1a to the lower edge of the image D mp imaging value, contour O v lower edge image width W mp imaging value, image width at the widest point of contour O 1 W 1mp The imaging value and its light intensity imaging value Y 1mp , the contour O 1r span R mp imaging value and its light intensity Y rmp imaging value, that is, the imaging detection value group M p (D mp , W mp , W 1mp , Y 1mp , R mp , Y rmp ) ; ( 2a) The imaging value group J ep [ ( D t , D tp ), (W v , W vp ), (W w =W vp /4, W wp =W vp /4), {(W 1 =W vp /16, W 1p =W vp /16), (Y 1 , Y 1p )}, {(Y h , Y hp ), (R h , R hp )}, {(Y 1 , Y 1p ), (R 1 , R 1p )}] comparison to determine traffic volume category;
所述判断交通体类别的判定规则如下,(1)车辆与行人判定:当指定交通体的宽度检测值W
m,(1.1)不小于交通体数字设定判断值车辆宽度W
v时,该交通体判定为车辆;否则,(1.2)小于的但不小于相对应的行人宽度设定判断值W
w时,则判定为行人或单车类;否则,(1.3)或更小的但不小于车灯直径/宽度设定判断值W
1,并且其光强检测值不小于车灯光强等级设定判断值Y
1,则判定为车灯,进而判定为车辆或电单车;(2)车灯光判定:在强阴/极暗环境光强下,当指定交通体轮廓区光强检测值Y
m,(2.1)不小于交通体数字设定判断值车远光灯光强等级Y
h,同时其区域纵向跨度检测值R
m不小于相应的车远光灯光区跨度设定判断值Rb,或者,其跨度包括了远光灯跨度判断设定区--环境指定区域,临时判断为远光灯,当该临时判定持续了若干连续帧图像后,则判定为车远光灯;该连续帧图像数所用时长称为判定远光灯需用时长,大于估计常规开启远光灯最小时长,本实施例将其设定为1秒,等同于25帧PAL制视频图像时间;否则,(2.2)较小的Y
m但不小于相应的车近光灯光强等级设定判断值Y
11,或其R
m不小于相应的车近光灯光区跨度设定判断值R
11,则判定为车近光灯;
The judging rules for judging the category of the traffic body are as follows: (1) vehicle and pedestrian judgment: when the width detection value W m of the specified traffic body, (1.1) is not less than the traffic body digital setting judgment value vehicle width W v , the traffic body Otherwise, if (1.2) is smaller than but not less than the corresponding pedestrian width setting judgment value W w , then it is judged as a pedestrian or a bicycle; otherwise, (1.3) or smaller but not less than the vehicle light The diameter/width setting judgment value W 1 , and its light intensity detection value is not less than the car light intensity level setting judgment value Y 1 , then it is judged as a car light, and then it is judged as a vehicle or a motorcycle; (2) Car light judgment : In strong overcast/dark ambient light intensity, when the light intensity detection value Y m of the designated traffic body outline area is not less than the traffic body digital setting judgment value Y h , and the area vertically The span detection value R m is not less than the corresponding vehicle high beam light area span setting judgment value Rb, or its span includes the high beam span judgment setting area - the environment designated area, temporarily judged as high beam, when the After the provisional determination lasts for several consecutive frames of images, it is determined to be the high beam of the vehicle; the time required for the number of consecutive frames of images is called the time required for judging the high beam, which is greater than the estimated minimum time for turning on the high beam conventionally. Set it to 1 second, which is equivalent to 25 frames of PAL video image time; otherwise, (2.2) the smaller Y m but not less than the corresponding car low beam light intensity level setting judgment value Y 11 , or its R m is not less than If the judgment value R 11 is set for the span of the corresponding vehicle low beam light area, it is determined to be the vehicle low beam light;
S5e,根据获取的车前信息环境光强检测值Y
em的光强等级,交通体信息检测得到的车辆行人与车灯光,安全灯控逻辑产生控制车灯信号;所述安全灯控逻辑包括:
S5e, according to the light intensity level of the obtained vehicle front information environmental light intensity detection value Y em , the vehicle pedestrian and vehicle lights detected by the traffic body information, the safety light control logic generates a control light signal; the safety light control logic includes:
1)当环境光强为白天时,关闭远光灯、关闭近光灯,关闭示宽灯;1) When the ambient light intensity is daytime, turn off the high beam, turn off the low beam, and turn off the width light;
2)当环境光强为强阴时,开启近光灯和示宽灯;2) When the ambient light intensity is strong and cloudy, turn on the dipped beam and width indicator lights;
3)当环境光强为极暗时,开启远光灯;3) When the ambient light intensity is extremely dark, turn on the high beam;
4)当环境光强为强阴或极暗,交通体为远光灯时,开启闪烁远光灯;否则,4) When the ambient light intensity is overcast or extremely dark and the traffic body is high beam, turn on and flash the high beam; otherwise,
5)当环境光强为强阴或极暗,交通体为车辆或行人时,关闭远光灯,保持/开启其它灯;5) When the ambient light intensity is overcast or extremely dark, and the traffic body is vehicles or pedestrians, turn off the high beam and keep/turn on other lights;
所述闪烁包括闪烁次数,单次时长,和间隔时长,本实施例设定其次数为3,单次时长120毫秒,等同于3帧图像(PAL标准25帧/秒),远小于上述判定远光灯需用时长1000毫秒,间隔时长为320毫秒,合计每次开启闪烁用灯时长1秒钟。The flicker includes the number of flickers, a single duration, and an interval duration. In this embodiment, the number of times is set to 3, and the single duration is 120 milliseconds, which is equivalent to 3 frames of images (PAL standard 25 frames/second), far less than the above-mentioned judgment The light takes 1000 milliseconds, and the interval is 320 milliseconds. The total time for each flashing light is 1 second.
Claims (19)
- 一种车灯智能控制***,其特征在于包括,车前视频分析子***,安全灯控逻辑模块,通信子***,车灯控制装置;所述车前视频分析子***产生指定车前信息,包括指定车前环境信息或与指定车前交通体信息,所述指定车前信息经过安全灯控逻辑模块后,产生控制车灯信号,通过通信子***连接到车灯控制装置,控制车灯;An intelligent control system for car lights, characterized in that it includes a front-of-car video analysis subsystem, a safety light control logic module, a communication subsystem, and a car light control device; the front-of-car video analysis subsystem generates specified front-of-car information, including Designate the environment information in front of the vehicle or the traffic body information in front of the designated vehicle. After the designated front information passes through the safety light control logic module, it generates a signal to control the vehicle light, and connects to the vehicle light control device through the communication subsystem to control the vehicle light;所述指定车前环境信息包括指定环境光强类别:不需要开启车灯的白天/需要开启车远光灯的极暗/需要开启车灯的强阴;所述指定车前交通体信息包括交通体类别:车辆/行人包括单车/车远光灯。The specified environmental information in front of the vehicle includes the specified environmental light intensity category: daytime when the lights do not need to be turned on/very dark when the high beam lights need to be turned on/strong shade when the lights need to be turned on; the specified traffic body information in front of the vehicle includes traffic Body category: vehicle/pedestrian including bicycle/vehicle high beam.
- 根据权利要求1所述车灯智能控制***,其特征在于,所述通信子***指无线或/或与有线通信子***,包括CAN或/或与蓝牙或/或与WIFI/或与ModBus或/或与ZigBee。According to claim 1, the vehicle light intelligent control system is characterized in that the communication subsystem refers to a wireless or/or wired communication subsystem, including CAN or/or with Bluetooth or/or with WIFI/or with ModBus or/ Or with ZigBee.
- 根据权利要求1所述车灯智能控制***,其特征在于,所述安全灯控逻辑模块,嵌入在图像处理子***或在通信子***或在车灯控制装置之中,用硬件或/或与软件实现。The vehicle light intelligent control system according to claim 1, characterized in that, the safety light control logic module is embedded in the image processing subsystem or in the communication subsystem or in the vehicle light control device, using hardware or/or in conjunction with Software Implementation.
- 根据权利要求1所述车灯智能控制***,其特征在于,所述控制车灯信号包括开启或关闭或闪烁如下车灯及其组合,近光灯,远光灯,示宽灯。According to the vehicle light intelligent control system according to claim 1, characterized in that, said control vehicle light signal includes turning on or off or flashing the following vehicle lights and combinations thereof, low beam light, high beam light and width indicator light.
- 根据权利要求1所述车灯智能控制***,其特征在于,所述车灯控制装置包括原人工控制车灯优先:临时优先和临时切断;临时优先指只要人操作开启车灯就进入原人工控制车灯状态,该操作完成关闭车灯后,自动恢复本控制;临时切断指需要人工操作另外的切断按键和恢复或重启按键,才能切断和恢复本控制;两者用硬件电路或/或与嵌入式软件实现。The vehicle light intelligent control system according to claim 1, characterized in that the vehicle light control device includes the original manual control vehicle light priority: temporary priority and temporary cut-off; temporary priority means that as long as a human operates to turn on the vehicle light, it will enter the original manual control The state of the car light, after the operation is completed and the car light is turned off, the control is automatically restored; the temporary cut-off means that the other cut-off button and the restore or restart button need to be manually operated to cut off and restore the control; the two use hardware circuits or/or with embedded software implementation.
- 一种车灯智能控制方法,其特征在于包括,S1,配置检测基础参数;S2,获取车前视频图像;S3,从所述车前视频图像提取指定车前信息,包括环境信息;S4,计算与判断指定车前信息特征:计算所述指定环境信息光强并判断其光强等级类别;S5,安全灯控逻辑根据所述指定环境光强等级类别,产生控制车灯信号;An intelligent control method for vehicle lights, characterized by comprising: S1, configuring basic parameters for detection; S2, acquiring video images in front of the vehicle; S3, extracting specified information in front of the vehicle from the video images in front of the vehicle, including environmental information; S4, calculating and judging the characteristics of the designated front information of the vehicle: calculating the light intensity of the designated environment information and judging its light intensity level category; S5, the safety light control logic generates a control vehicle light signal according to the specified environment light intensity level category;所述配置检测基础参数包括标定提供所述车前视频的摄像机内外参数,设定环境光强等级类别的判断值,用以判断环境光强类别是白天/极暗/强阴。The configuration detection basic parameters include calibrating the internal and external parameters of the camera that provides the video in front of the car, and setting the judgment value of the ambient light intensity level category to determine whether the ambient light intensity category is daytime/very dark/strong shade.
- 根据权利要求6所述车灯智能控制方法,其特征在于,S1-1,所述环境光强等级类别设定分为:1.白天,不需要开启车灯的光强;2.极暗,包括夜间无路灯或弱路灯需要开启远光车灯的光强;3.强阴,光强于极暗,包括黄昏/清晨/夜间强路灯/极低云层阴天/隧道内需要开启车灯的光强;各环境光强等级类别或用设定的各对应灰度值作为其判断值;或还设定指定时间间隔的该类别之间灰度值之差最大值,作为判断是否属于环境光强变化差的判断值,小于该值的判定环境光强差,称为环境之差判断值。According to the intelligent control method for vehicle lights according to claim 6, characterized in that, S1-1, the setting of the environmental light intensity level category is divided into: 1. During the day, the light intensity of the vehicle lights does not need to be turned on; 2. Extremely dark, Including the light intensity of no street lights or weak street lights at night that need to turn on the high beam lights; 3. Strong overcast, light intensity is stronger than extremely dark, including dusk/early morning/night strong street lights/very low cloud cloudy days/tunnels that need to turn on lights Light intensity; each category of ambient light intensity level or use the set corresponding gray value as its judgment value; or also set the maximum value of the gray value difference between the categories at a specified time interval as a judgment whether it belongs to ambient light The judging value of the strong change difference, and the judging ambient light intensity difference smaller than this value is called the judging value of the environmental difference.
- 根据权利要求6所述车灯智能控制方法,其特征在于,S3-1,所述提取指定环境信息的方法指任何提取图像目标背景的算法,包括直接提取图像中所有像素的灰度值,或提取图像中环境指定区域所有像素的灰度值。According to the intelligent control method of vehicle lights according to claim 6, it is characterized in that, S3-1, the method of extracting specified environmental information refers to any algorithm for extracting the target background of the image, including directly extracting the gray value of all pixels in the image, or Extracts the grayscale values of all pixels in an environment-specified region of the image.
- 根据权利要求6所述车灯智能控制方法,其特征在于,S4-1,所述计算指定环境信息光强包括计算与所提取的指定环境信息即所述灰度值相关的数值,包括其平均值,作为该环境光强特征的检测值。According to the intelligent control method of vehicle lights according to claim 6, characterized in that, S4-1, said calculating the light intensity of specified environmental information includes calculating the values related to the extracted specified environmental information, that is, the gray value, including the average value, as the detection value of the ambient light intensity feature.
- 根据权利要求6所述车灯智能控制方法,其特征在于,S4-2,所述判断指定环境光强等级类别是将所述指定环境光强检测值与所述设定环境光强等级类别判断值对比,其规则包括:当指定环境光强检测值,(1)不小于设定环境光强等级类别白天判断值时,该环境光强判定为白天;(2)不大于相应极暗判断值时,该环境光强判定为极暗;(3)介于极暗判断值与白天判断值之间时,该环境光强判定为强阴。According to the intelligent control method of vehicle lights according to claim 6, characterized in that, in S4-2, said judging the category of the specified ambient light intensity level is to judge the specified ambient light intensity detection value and the set ambient light intensity level category Value comparison, the rules include: when the specified ambient light intensity detection value, (1) is not less than the daytime judgment value of the set ambient light intensity level category, the ambient light intensity is judged as daytime; (2) is not greater than the corresponding extremely dark judgment value , the ambient light intensity is judged as extremely dark; (3) when it is between the extremely dark judgment value and the daytime judgment value, the ambient light intensity is judged as strong overcast.
- 根据权利要求10所述车灯智能控制方法,其特征在于,S4-3,所述判断指定环境光强等级类别是将所述指定环境光强检测值与所述设定环境光强等级类别判断值对比时,参考其之前图像相关光强判定及其与该之前图像相关光强灰度值之差,其规则包括在判断上述规则(1)之前先判断:当指定环境光强检测值,(0)大于其之前图像相关光强类别,其与该之前图像相关光强灰度值之差还不小于环境之差判断值,该环境光强被判定为临时光强;否则,执行上述规则(1)。According to the intelligent control method of vehicle lights according to claim 10, characterized in that, in S4-3, said judging the category of the specified ambient light intensity level is to judge the specified ambient light intensity detection value and the set ambient light intensity level category When comparing values, refer to the previous image-related light intensity judgment and the difference between the previous image-related light intensity gray value. The rules include judging before the above-mentioned rule (1): When specifying the ambient light intensity detection value, ( 0) is greater than its previous image-related light intensity category, and the difference between it and the previous image-related light intensity gray value is not less than the environment difference judgment value, and the environment light intensity is judged as temporary light intensity; otherwise, execute the above rules ( 1).
- 根据权利要求6所述车灯智能控制方法,其特征在于,S5-1,所述安全灯控逻辑包括:According to the intelligent control method of vehicle lights according to claim 6, characterized in that, S5-1, the safety light control logic includes:1)当环境光强为白天时,关闭远光灯,近光灯,示宽灯;1) When the ambient light intensity is daytime, turn off the high beam, low beam and width indicators;2)当环境光强为强阴时,开启/保持近光灯和示宽灯;2) When the ambient light intensity is strong and cloudy, turn on/keep the dipped beam and width indicators;3)当环境光强为极暗时,开启/保持远光灯。3) When the ambient light intensity is extremely dark, turn on/maintain the high beam.
- 根据权利要求6所述车灯智能控制方法,其特征在于,S1a,所述S1中配置检测基础参数还包括设定交通体类别判断值,用以判断交通体类别是车辆/行人/远光灯;S3a,所述S3中从所述车前视频提取的指定车前信息还包括交通体信息;S4a,所述S4中计算与判断指定车前信息特征还包括计算指定交通体信息特征值并判断其类别;S5a,所述S5中安全灯控逻辑还包括根据所述指定车前交通体类别产生控制车灯信号。According to the intelligent control method for vehicle lights according to claim 6, it is characterized in that, S1a, configuring the detection basic parameters in S1 also includes setting the judgment value of the traffic body type to judge whether the traffic body type is vehicle/pedestrian/high beam S3a, the information in front of the specified vehicle extracted from the video in front of the vehicle in S3 also includes traffic body information; S4a, calculating and judging the features of the information in front of the vehicle in the S4 also includes calculating the feature value of the specified traffic body information and judging Its category; S5a, the safety light control logic in S5 also includes generating a signal to control the vehicle light according to the category of the designated traffic body in front of the vehicle.
- 根据权利要求13所述车灯智能控制方法,其特征在于,S1a-1,所述设定交通体类别判断值是一组估计最小可辨识交通体类别的特征值,包括特征宽度/直径判断值,特征光强等级及其光照区纵向跨度判断值;1.车辆宽度:设定以常规最小车辆宽度,并设定其数倍于行人宽度;2.行人宽度:设定以常规最小行人宽度,并设定电单车/摩托车与其同宽度;3.车灯直径:设定以常规最小车灯直径,还设定其与电单车或摩托车灯相同,其光强设定以车灯最小光强;4.车远光灯光强等级:设定以强于车近光灯光强,其光区最小纵向/横向跨度以近光灯光区常规最大纵向/横向跨度设定,或者,以设定区域作为远光灯跨度判断设定;5.车近光灯光强等级:设定以强于车灯光强,其光区纵向/横向跨度设定常规最大和最小值;这些判断值可以根据使用情况调整再优化。According to the intelligent control method of vehicle lights according to claim 13, it is characterized in that, S1a-1, the set traffic body type judgment value is a set of characteristic values of the estimated minimum identifiable traffic body type, including the characteristic width/diameter judgment value , the characteristic light intensity level and the judging value of the longitudinal span of the illuminated area; 1. Vehicle width: set the conventional minimum vehicle width, and set it several times the pedestrian width; 2. Pedestrian width: set the conventional minimum pedestrian width, And set the motorcycle/motorcycle with the same width; 3. Car light diameter: set it to the conventional minimum car light diameter, and also set it to be the same as the motorcycle or motorcycle light, and set the light intensity to the minimum light of the car light Strong; 4. Car high beam light intensity level: set to be stronger than the car low beam light intensity, and the minimum vertical/horizontal span of the light area is set by the conventional maximum vertical/horizontal span of the low beam light area, or the set area is used as the High beam span judgment setting; 5. Car low beam light intensity level: set to be stronger than the car light intensity, and the vertical/horizontal span of the light zone is set to the conventional maximum and minimum values; these judgment values can be adjusted according to the usage Re-optimize.
- 根据权利要求13所述车灯智能控制方法,其特征在于,S3a-1,从所述车前视频提取指定车前交通体信息方法包括用图像目标检测算法直接从图像中或从图像中的交通体指定区域,提取交通体轮廓,获取该轮廓的高、宽或直径尺寸和代表其光强的灰度值;所述图像目标检测算法包括R-CNN算法系列或YOLO算法系列,及灰度对比,但不限于这些。According to claim 13, the vehicle light intelligent control method is characterized in that, S3a-1, the method of extracting the specified traffic body information in front of the vehicle from the video in front of the vehicle includes using the image target detection algorithm to directly extract the information from the image or from the traffic in the image The designated area of the body, extract the contour of the traffic body, obtain the height, width or diameter of the contour and the gray value representing its light intensity; the image target detection algorithm includes R-CNN algorithm series or YOLO algorithm series, and grayscale comparison , but not limited to these.
- 根据权利要求13所述车灯智能控制方法,其特征在于,S4a-1,所述计算指定交通体信息特征值包括,提取所述指定交通体轮廓宽度或直径作为其检测值;计算所述指定交通体轮廓光强灰度值的平均值及其纵向/横向跨度作为其检测值。According to the intelligent control method of vehicle lights according to claim 13, characterized in that, S4a-1, said calculating the characteristic value of the specified traffic body information includes extracting the specified traffic body contour width or diameter as its detection value; calculating the specified The average value of the gray value of the light intensity of the traffic body contour and its vertical/horizontal span are taken as its detection value.
- 根据权利要求13所述车灯智能控制方法,其特征在于,S4a-2,所述判断指定交通体类别包括判断车辆/行人和判断车灯光;将所述交通体轮廓宽度或直径检测值,与所述设定交通体宽度判断值对比,符合交通体判定规则的图像目标,判定为相应交通体:车辆或车灯(也是车辆)或行人或单车;将所述交通体轮廓光强检测值及其纵向/横向跨度检测值,与所述交通体光强等级设定判断值和纵向/横向跨度设定判断值综合对比,符合交通体判定规则的图像目标,判定为相应车辆灯光:车远光灯或车近光灯/车灯。According to the intelligent control method of vehicle lights according to claim 13, it is characterized in that, S4a-2, said judging the designated traffic body category includes judging vehicles/pedestrians and judging car lights; combining the detected value of the traffic body contour width or diameter with The comparison of the set traffic body width judgment value, the image target meeting the traffic body judgment rules, is judged as the corresponding traffic body: vehicle or car light (also vehicle) or pedestrian or bicycle; the described traffic body contour light intensity detection value and The detected value of the vertical/horizontal span is comprehensively compared with the set judgment value of the light intensity level of the traffic body and the set judgment value of the vertical/horizontal span. The image object conforming to the traffic body judgment rule is judged to be the corresponding vehicle light: car high beam headlights or car dipped beam/headlights.
- 根据权利要求17所述车灯智能控制方法,其特征在于,S4a-3,所述判断指定交通体类别的判定规则如下:(1)车辆与行人判定:当指定交通体轮廓宽度检测值,(1.1)不小于交通体车辆宽度设定判断值时,该交通体判定为车辆;否则,(1.2)较小的该宽度检测值但不小于相关行人宽度设定判断值时,则判定为行人或单车类;否则,(1.3)或更小的该宽度检测值但不小于相关车灯直径/宽度设定判断值,并且在强阴/极暗环境光强下其光强检测值不小于相关车灯光强设定判断值,则判定为车灯,进而判为车辆或电单车;(2)车灯光判定:在强阴/极暗环境光强下,当指定交通体轮廓区域光强检测值,(2.1)不小于交通体的车远光灯光强设定判断值,同时其区域纵向/横向跨度检测值不小于相关车远光灯光区跨度设定判断值,或者,其区域跨度包括了远光灯跨度判断设定区,则判定为车远光灯;否则,(2.2)较小的该轮廓区光强检测值但不小于相关车近光灯光强设定判断值,或其区域纵向跨度检测值不小于相关车近光灯光区跨度设定判断值,则判定为车近光灯;According to the intelligent control method of vehicle lights according to claim 17, it is characterized in that, S4a-3, the determination rules for determining the category of the specified traffic body are as follows: (1) Vehicle and pedestrian determination: when the detected value of the contour width of the specified traffic body, ( 1.1) When the vehicle width of the traffic body is not smaller than the set judgment value of the vehicle width, the traffic body is judged as a vehicle; otherwise, (1.2) when the width detection value is smaller but not less than the judgment value of the width of the relevant pedestrian, it is judged as a pedestrian or Bicycles; otherwise, the detection value of the width (1.3) or less but not less than the judgment value of the diameter/width setting of the relevant car lights, and the detection value of the light intensity in the strong shade/dark environment is not less than that of the relevant vehicle If the light intensity is set to the judgment value, it is judged to be a car light, and then judged to be a vehicle or a motorcycle; (2) Car light judgment: under strong cloudy/very dark ambient light intensity, when the light intensity detection value of the specified traffic contour area , (2.1) Not less than the judging value of the vehicle high beam light intensity setting of the traffic body, and the detection value of its regional longitudinal/horizontal span is not less than the setting judging value of the relevant vehicle high beam light area span, or its regional span includes far If it is not within the span judgment setting area of the light, it is judged as the high beam of the vehicle; otherwise, (2.2) the light intensity detection value of the contour area is smaller but not less than the judgment value of the low beam light intensity setting of the relevant car, or the vertical span of the area If the detection value is not less than the judgment value set for the span of the low beam light area of the relevant vehicle, it is determined to be the low beam light of the vehicle;或上述步骤(2.1)车远光灯的判定还包括如下步骤(2.1a):将(2.1)对车远光灯的判定作为临时判断,将持续了若干连续帧图像的所述临时判断作为远光灯判定;该连续帧图像数所用时长称为判定远光灯需用时长,设定大于估计常规开启远光灯最小时长。Or the judgment of the above-mentioned steps (2.1) the high beam of the car also includes the following steps (2.1a): (2.1) is used as a temporary judgment for the judgment of the high beam of the car, and the temporary judgment that has continued several continuous frame images is taken as a far Judgment of light; the time required for the number of consecutive frames of images is called the time required for judging the high beam, and the setting is greater than the estimated minimum time for turning on the high beam.
- 根据权利要求13所述车灯智能控制方法,其特征在于,S5a-1,所述安全灯控逻辑还包括:4)当环境光强是强阴或极暗,有车辆远光灯时,或启动闪烁远光灯;According to the intelligent control method of vehicle lights according to claim 13, it is characterized in that, S5a-1, the safety light control logic further includes: 4) when the ambient light intensity is strong shade or very dark, and there is a vehicle high beam, or Start flashing high beams;5)当环境光强是强阴或极暗,有车辆或行人时,关闭远光灯,保持/开启其它灯;5) When the ambient light intensity is overcast or extremely dark, and there are vehicles or pedestrians, turn off the high beam and keep/turn on other lights;所述闪烁包括闪烁次数,间隔和单次时长,其单次时长小于判定远光灯需用时长。The flickering includes the number of flickers, an interval and a single duration, and the single duration is shorter than the time required for judging the high beam.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2022/000029 WO2023164786A1 (en) | 2022-03-01 | 2022-03-01 | Intelligent control system and method for vehicle lamp |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2022/000029 WO2023164786A1 (en) | 2022-03-01 | 2022-03-01 | Intelligent control system and method for vehicle lamp |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2023164786A1 true WO2023164786A1 (en) | 2023-09-07 |
Family
ID=87882737
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2022/000029 WO2023164786A1 (en) | 2022-03-01 | 2022-03-01 | Intelligent control system and method for vehicle lamp |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2023164786A1 (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005059814A (en) * | 2003-08-20 | 2005-03-10 | Ichikoh Ind Ltd | Lighting system for vehicle |
US20160016503A1 (en) * | 2012-11-16 | 2016-01-21 | Zhengzhou Yutong Bus Co., Ltd. | Intelligent control method and system for automobile headlights |
JP2017094965A (en) * | 2015-11-25 | 2017-06-01 | 富士重工業株式会社 | Vehicle exterior environment recognition device |
WO2017186118A1 (en) * | 2016-04-27 | 2017-11-02 | 李勇 | Profile-based intelligent automotive light controller and control method |
US20210046861A1 (en) * | 2019-08-12 | 2021-02-18 | Nvidia Corporation | Automatic high beam control for autonomous machine applications |
WO2021193702A1 (en) * | 2020-03-27 | 2021-09-30 | 株式会社小糸製作所 | Vehicle lamp system, light distribution control device, light distribution control method, vehicle determination device, and vehicle determination method |
-
2022
- 2022-03-01 WO PCT/CN2022/000029 patent/WO2023164786A1/en unknown
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005059814A (en) * | 2003-08-20 | 2005-03-10 | Ichikoh Ind Ltd | Lighting system for vehicle |
US20160016503A1 (en) * | 2012-11-16 | 2016-01-21 | Zhengzhou Yutong Bus Co., Ltd. | Intelligent control method and system for automobile headlights |
JP2017094965A (en) * | 2015-11-25 | 2017-06-01 | 富士重工業株式会社 | Vehicle exterior environment recognition device |
WO2017186118A1 (en) * | 2016-04-27 | 2017-11-02 | 李勇 | Profile-based intelligent automotive light controller and control method |
US20210046861A1 (en) * | 2019-08-12 | 2021-02-18 | Nvidia Corporation | Automatic high beam control for autonomous machine applications |
WO2021193702A1 (en) * | 2020-03-27 | 2021-09-30 | 株式会社小糸製作所 | Vehicle lamp system, light distribution control device, light distribution control method, vehicle determination device, and vehicle determination method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109152185A (en) | A kind of multi-source perception intelligent street lamp control system | |
CN106218492B (en) | A kind of night running light automatic switching control equipment and switching method | |
CN107972569B (en) | Vehicle lamp set control method, device and system and vehicle | |
CN109194873B (en) | Image processing method and device | |
JP7268001B2 (en) | Arithmetic processing unit, object identification system, learning method, automobile, vehicle lamp | |
CN105730321B (en) | A kind of vehicular high beam light intelligent control system and control method | |
CN109703460B (en) | Multi-camera complex scene self-adaptive vehicle collision early warning device and early warning method | |
JP4853437B2 (en) | Vehicle perimeter monitoring system | |
US11325522B2 (en) | Automatic light system | |
CN106295494B (en) | Vehicle exterior environment recognition device | |
CN208479822U (en) | A kind of automobile-used panoramic looking-around system | |
CN108650495A (en) | A kind of automobile-used panoramic looking-around system and its adaptive light compensation method | |
JP2009017157A (en) | Image processor, method and program | |
CN106915295A (en) | The control method and device of automobile front lamp state | |
CN106778534A (en) | Surrounding environment method for recognition of lamplight in a kind of vehicle traveling | |
JP2016196233A (en) | Road sign recognizing device for vehicle | |
CN112001208A (en) | Target detection method and device for vehicle blind area and electronic equipment | |
CN113438779A (en) | Method for triggering street lamp illumination intensity adjustment based on image | |
JP2010188826A (en) | Display device for vehicle | |
CN104008518B (en) | Body detection device | |
JP2003196786A (en) | Method and device for detecting moving object | |
CN111688568B (en) | Brightness detection method, vehicle lamp control method, system thereof and storage medium | |
WO2023164786A1 (en) | Intelligent control system and method for vehicle lamp | |
KR101511586B1 (en) | Apparatus and method for controlling vehicle by detection of tunnel | |
KR20180093248A (en) | Method and apparatus for controlling head lamp of vehicle |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22929240 Country of ref document: EP Kind code of ref document: A1 |