Obstacle identification method and device, equipment and storage medium
By obtaining vehicle driving data and obstacle motion parameters, combining the fusion perception of monocular cameras and millimeter wave radars, we can determine whether there is a collision risk between vehicles and obstacles, and solve the problem of poor accuracy and stability of obstacle identification in the prior art. Improves the vehicle's obstacle detection capabilities in complex environments.
Patent Information
- Application Number
- CN202410893643.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-04
- Publication Date
- 2024-11-26
AI Technical Summary
The existing obstacle recognition methods cannot effectively identify target obstacles in complex traffic scenarios, and the input data comes from a single sensor, which lacks the assistance of visual sensors, resulting in poor recognition accuracy and stability.
By obtaining vehicle driving data, it determines its estimated driving trajectory, and determines the estimated moving trajectory of the obstacle based on the motion parameters of the obstacles in the target area of the vehicle. It determines whether there is a collision risk based on the trajectory of the vehicle and the obstacle. If there is, determines the obstacle. Is the object an obstacle to the target? This method uses the fusion perception of a monocular camera and a millimeter-wave radar, which enhances the detection reliability and accuracy of target obstacles.
It significantly improves the reliability and accuracy of target obstacle detection in complex environments, optimizes the comprehensive efficiency of the perceived fusion algorithm, and enhances the safety and reliability of the vehicle longitudinal control function.
Smart Images

Figure CN119024325A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of assisted driving of intelligent vehicles, and in particular to an obstacle recognition method, device, equipment and storage medium. Background Art
[0002] In recent years, with the continuous advancement of science and technology, people's demand for travel safety and convenience has continued to increase. In this context, the L2-level intelligent driving assistance function for vehicles is of great significance in alleviating the workload of drivers, improving traffic safety, and increasing transportation efficiency, and has become a key direction for the development of the industry.
[0003] At present, in the L2 driving assistance control of vehicles, higher requirements are put forward for the recognition of obstacles in complex traffic scenes by the perception module. During the driving process of the vehicle, obstacles usually have the characteristics of small reflection volume, unclear visual features, and complex movement trajectories. Therefore, stable recognition and tracking of target obstacles have become an important factor affecting the longitudinal control function of the vehicle. Stable and accurate recognition and output of target obstacles are of great significance to the safety and reliability of commercial vehicle driving assistance functions.
[0004] The disadvantage of the existing obstacle recognition method is that it does not use traffic scene information to classify the movement state of obstacles, and the input data often comes from a single sensor, lacking the assistance of visual sensors, which leads to the inability to effectively identify the target obstacles. Therefore, the present invention proposes an obstacle recognition method that can effectively improve the output stability and accuracy of target obstacles in complex scenes and improve the overall performance of the perception fusion algorithm. Summary of the invention
[0005] The present application provides an obstacle recognition method, device, equipment and storage medium to solve the problems of poor accuracy and stability of input signals during vehicle driving, inability to effectively identify target obstacles and estimate collision risks, etc.
[0006] In a first aspect, the present application provides an obstacle recognition method, the method comprising:
[0007] When it is determined by the first detection device that there is an obstacle in the target area, obtaining driving data of the vehicle, and determining an estimated driving trajectory of the vehicle based on the driving data;
[0008] Acquire motion parameters of the obstacle through the first detection device, and determine an estimated motion trajectory of the obstacle based on the motion parameters;
[0009] Determining whether the vehicle has a collision risk according to the estimated driving trajectory and the estimated motion trajectory;
[0010] If there is a collision risk with the vehicle, it is determined whether the obstacle is a target obstacle.
[0011] Optionally, the driving data includes: driving speed and yaw angular velocity, and determining the estimated driving trajectory of the vehicle based on the driving data includes:
[0012] Determining a driving radius of the vehicle according to the driving speed and the yaw angular velocity;
[0013] An estimated driving trajectory of the vehicle is determined according to the driving radius.
[0014] Optionally, there are multiple obstacles, and the motion parameters include: position, motion speed, and orientation angle. The step of determining the estimated motion trajectory of the obstacle based on the motion parameters includes:
[0015] For any one obstacle among the multiple obstacles, an estimated motion trajectory of the obstacle is determined according to the position, motion speed and orientation angle.
[0016] Optionally, determining whether the vehicle has a collision risk according to the estimated driving trajectory and the estimated motion trajectory includes:
[0017] Determine the motion state of each obstacle according to the estimated motion trajectories of the multiple obstacles, wherein the motion state includes: traversing motion, same-direction motion, and stationary motion;
[0018] According to the motion states of multiple obstacles, the collision risk area of each obstacle is determined;
[0019] Determine whether there is a target collision risk area in the estimated driving trajectory among the multiple collision risk areas;
[0020] If so, it is determined that the vehicle has a collision risk.
[0021] Optionally, determining the collision risk area of each obstacle according to the motion states of the multiple obstacles includes:
[0022] For an obstacle whose motion state is lateral motion or same-direction motion, determining a collision risk area of the obstacle according to the estimated motion trajectory of the obstacle and the estimated driving trajectory;
[0023] For an obstacle whose motion state is stationary, the position of the obstacle is used as the collision risk area of the obstacle.
[0024] Optionally, if the vehicle has a collision risk, determining whether the obstacle is a target obstacle includes:
[0025] Determine a candidate obstacle from obstacles corresponding to a plurality of target collision risk areas;
[0026] Detecting the real-time motion trajectory of the candidate obstacle in real time by a second detection device;
[0027] If the real-time motion trajectory of the candidate obstacle coincides with the estimated motion trajectory within a preset time period, the candidate obstacle is determined to be the target obstacle.
[0028] Optionally, determining a candidate obstacle from obstacles corresponding to a plurality of target collision risk areas includes:
[0029] Determine a plurality of obstacles corresponding to a plurality of target collision risk areas, and determine a collision moment of each obstacle according to the estimated running trajectories of the plurality of obstacles and the estimated driving trajectory of the vehicle;
[0030] sorting the obstacles in ascending order of the multiple collision moments, wherein the collision risk value of the obstacle in the front order is higher than the collision risk value of the obstacle in the back order;
[0031] The obstacles whose serial numbers meet a preset condition are taken as the candidate obstacles, wherein the preset condition includes: the serial number is less than or equal to a preset serial number.
[0032] In a second aspect, the present application provides an obstacle recognition device, the device comprising:
[0033] An acquisition module, used for acquiring the driving data of the vehicle when it is determined by the first detection device that there is an obstacle in the target area;
[0034] A determination module, configured to determine an estimated driving trajectory of the vehicle based on the driving data;
[0035] The acquisition module is further used to acquire the motion parameters of the obstacle through the first detection device;
[0036] The determination module is further used to determine the estimated motion trajectory of the obstacle based on the motion parameters;
[0037] A judgment module, for judging whether the vehicle has a collision risk according to the estimated driving trajectory and the estimated motion trajectory;
[0038] The determination module is further configured to determine whether the obstacle is a target obstacle when there is a collision risk with the vehicle.
[0039] Optionally, the driving data includes: driving speed and yaw angular velocity;
[0040] The determination module is further used to determine the driving radius of the vehicle according to the driving speed and the yaw angular velocity; and to determine the estimated driving trajectory of the vehicle according to the driving radius.
[0041] Optionally, there are multiple obstacles, and the motion parameters include: position, motion speed and orientation angle;
[0042] The determination module is further configured to determine, for any one of the multiple obstacles, an estimated motion trajectory of the obstacle according to the position, motion speed, and orientation angle.
[0043] Optionally, the determination module is further used to determine the motion state of each obstacle according to the estimated motion trajectories of the multiple obstacles, the motion state including: traversing motion, same-direction motion and stationary; and determine the collision risk area of each obstacle according to the motion states of the multiple obstacles;
[0044] The judgment module is further used to judge whether there is a target collision risk area in the estimated driving trajectory among the multiple collision risk areas;
[0045] The determination module is further configured to determine that the vehicle has a collision risk when there is a target collision risk area located in the estimated driving trajectory among the multiple collision risk areas.
[0046] Optionally, the device further comprises a processing module;
[0047] The determination module is further configured to determine, for an obstacle whose motion state is a traversing motion or a co-directional motion, a collision risk area of the obstacle according to the estimated motion trajectory of the obstacle and the estimated driving trajectory;
[0048] The processing module is used for taking the position of an obstacle whose motion state is stationary as a collision risk area of the obstacle.
[0049] Optionally, the determination module is further used to determine a candidate obstacle from obstacles corresponding to multiple target collision risk areas;
[0050] The processing module is further used to detect the real-time motion trajectory of the candidate obstacle in real time through a second detection device;
[0051] The determination module is further configured to determine that the candidate obstacle is the target obstacle if the real-time motion trajectory of the candidate obstacle coincides with the estimated motion trajectory within a preset time period.
[0052] Optionally, the determination module is further used to determine multiple obstacles corresponding to the multiple target collision risk areas, and determine the collision moment of each obstacle according to the estimated running trajectories of the multiple obstacles and the estimated driving trajectory of the vehicle;
[0053] The processing module is further used to sort the obstacles according to the multiple collision moments in ascending order, wherein the collision risk value of the obstacle in the front arrangement order is higher than the collision risk value of the obstacle in the back arrangement order; and to take the obstacle whose serial number meets the preset condition as the candidate obstacle, wherein the preset condition includes: the serial number is less than or equal to the preset serial number.
[0054] In a third aspect, the present application provides an obstacle recognition device, comprising:
[0055] Memory;
[0056] processor;
[0057] Wherein, the memory stores computer-executable instructions;
[0058] The processor executes the computer-executable instructions stored in the memory to implement the obstacle recognition method as described in the first aspect and various possible implementations of the first aspect.
[0059] In a fourth aspect, the present application provides a computer storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the obstacle recognition method as described in the first aspect and various possible implementations of the first aspect.
[0060] The present application provides an obstacle recognition method, which obtains the vehicle driving data to determine its estimated driving trajectory, and then determines the estimated motion trajectory of the obstacle based on the motion parameters of the obstacle in the target area of the vehicle, and judges whether there is a collision risk between the vehicle and the obstacle according to the estimated driving trajectory of the vehicle and the estimated motion trajectory of the obstacle. If there is, it determines whether the obstacle is a target obstacle. This method significantly enhances the reliability and accuracy of target obstacle detection in complex environments during vehicle driving, thereby optimizing the comprehensive performance of the perception fusion algorithm. BRIEF DESCRIPTION OF THE DRAWINGS
[0061] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0062] Figure 1 A schematic diagram of a scenario of an obstacle recognition method provided in this application;
[0063] Figure 2 The process of an obstacle identification method provided for this application Figure 1 ;
[0064] Figure 3 The process of an obstacle identification method provided for this application Figure 2 ;
[0065] Figure 4 A schematic diagram of the structure of an obstacle recognition device provided in this application;
[0066] Figure 5 A schematic diagram of the structure of an obstacle recognition device provided in this application.
[0067] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION
[0068] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0069] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in sequences other than those illustrated or described herein.
[0070] In the embodiments of the present application, the words "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.
[0071] First, the terms involved in this application are explained:
[0072] Level 2 autonomous driving: Level 2 autonomous driving (Level 2 autonomous driving, L2 autonomous driving). Under the operating conditions specified by the system, the vehicle itself can control the steering and acceleration and deceleration of the car, and has adaptive cruise control and lane keeping assist functions.
[0073] Millimeter wave radar: Millimeter wave radar is a radar that works in the millimeter wave band (millimeter wave). Usually millimeter waves refer to the 30-300GHz frequency domain (wavelength is 1-10mm). The wavelength of millimeter waves is between microwaves and centimeter waves, so millimeter wave radar has some advantages of both microwave radar and photoelectric radar.
[0074] Compared with centimeter wave seekers, millimeter wave seekers are small in size, light in weight and have high spatial resolution. Compared with optical seekers such as infrared, laser, and television, millimeter wave seekers have strong ability to penetrate fog, smoke, and dust, and have the characteristics of all-weather (except for heavy rain) and all-day. In addition, the anti-interference and anti-stealth capabilities of millimeter wave seekers are also better than other microwave seekers. Millimeter wave radars can distinguish and identify very small targets, and can identify multiple targets at the same time; they have imaging capabilities, small size, good mobility and concealment.
[0075] Controller Area Network (CAN) communication: It is a serial communication network that effectively supports distributed control systems. It uses two lines, CAN high and CAN low, to achieve half-duplex communication. The CAN protocol follows the ISO / OSI model (ISO, International Organization for Standardization, full name International Organization for Standardization; OSI, OSI is the open system interconnection model defined by ISO in network communication, full name Open System Interconnection), especially the physical layer, data link layer and application layer. It adopts a multi-master working mode, there is no master-slave distinction between nodes, but there is a priority distinction between nodes, and the communication mode is flexible, which can realize point-to-point, point-to-multipoint and broadcast transmission of data without scheduling.
[0076] CAN communication uses a non-conflicting priority method for data transmission, and messages with higher priority are sent first. Each CAN frame consists of an identifier and a data field. The identifier is used to indicate the priority of the message and the identity of the sender, while the data field is used to carry the actual data information. The working process of CAN communication can be divided into two stages: message sending and message receiving. Various technologies and mechanisms are used to ensure the reliability and integrity of data transmission.
[0077] The CAN communication system is developed to meet the automotive industry's requirements for safety, comfort, convenience, low pollution and low cost. It supports high-speed communication, with a transmission rate of up to one megabit per second (1Mb / s, referring to the number of bits transmitted per second), and has 11-bit addressing and error detection capabilities. The characteristics of the CAN protocol include complete serial data communication, real-time support, high noise immunity and very low electromagnetic radiation.
[0078] With the continuous advancement of technology, people's demand for travel safety and convenience is increasing. In this context, the L2-level intelligent driving assistance function for vehicles is of great significance in alleviating the workload of drivers, improving traffic safety, and increasing transportation efficiency, and has become a key direction for the development of the industry.
[0079] At present, in the L2 driving assistance control of vehicles, higher requirements are put forward for the recognition of obstacles in complex traffic scenes by the perception module. During the driving process of the vehicle, obstacles usually have the characteristics of small reflection volume, unclear visual features, and complex movement trajectories. Therefore, stable recognition and tracking of target obstacles have become an important factor affecting the longitudinal control function of the vehicle. Stable and accurate recognition and output of target obstacles are of great significance to the safety and reliability of commercial vehicle driving assistance functions.
[0080] The existing obstacle recognition methods do not use traffic scene information to classify the movement state of obstacles, and the input data often comes from a single sensor and lacks the assistance of visual sensors, which leads to the inability to effectively identify target obstacles.
[0081] In response to the above problems, the present application provides an obstacle recognition method.
[0082] Figure 1 A schematic diagram of a scenario of an obstacle recognition method provided in this application. Figure 1 As shown, when the intelligent driving system of a vehicle performs obstacle recognition, for example, there is a vehicle 4 traveling forward and multiple obstacles, wherein obstacle 1 may be an obstacle that is about to move across, obstacle 2 is currently in a stationary state, and obstacle 3 may be an obstacle that is about to move in the same direction. In the above scenario, vehicle 4 needs to recognize obstacles according to the movement states of multiple obstacles.
[0083] This method obtains the vehicle's driving data to determine its estimated driving trajectory, and then determines the estimated motion trajectory of the obstacle based on the motion parameters of the obstacle in the vehicle's target area. According to the estimated driving trajectory of the vehicle and the estimated motion trajectory of the obstacle, it is determined whether there is a collision risk between the vehicle and the obstacle. If there is, it is determined whether the obstacle is a target obstacle. This method significantly enhances the reliability and accuracy of target obstacle detection in complex environments during vehicle driving, thereby optimizing the overall performance of the perception fusion algorithm.
[0084] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.
[0085] Figure 2 This is a process of an obstacle identification method provided in an embodiment of the present application. Figure 1 .like Figure 2 As shown, this embodiment provides an obstacle recognition method, including:
[0086] S101: When it is determined by a first detection device that there is an obstacle in a target area, the driving data of the vehicle is acquired, and based on the driving data, an estimated driving trajectory of the vehicle is determined.
[0087] Among them, the first detection device may, for example, refer to a monocular camera installed on the inside of the front windshield of a moving vehicle, and the target area may, for example, be the front of the vehicle, specifically, the distance limit range from a field of view angle less than 45° and a longitudinal distance exceeding 60 meters to the detectable distance of the first detection device.
[0088] When the vehicle is moving forward, when the first detection device, that is, the monocular camera, located on the inside of the front windshield of the vehicle detects that there is an obstacle in the target area, the vehicle's automatic driving system obtains the vehicle's driving data at this time and estimates the vehicle's driving trajectory based on the driving data.
[0089] Optionally, before the vehicle enters the driving state, it is necessary to power on the vehicle first so that the monocular camera and millimeter-wave radar installed on the inside of the vehicle's front windshield are powered, and then check whether the vehicle status, sensor calibration status and CAN communication are normal. If not, the vehicle's automatic driving system will give a fault prompt. If everything is normal, the obstacle recognition function enters the standby state.
[0090] S102: Acquire motion parameters of the obstacle through the first detection device, and determine an estimated motion trajectory of the obstacle based on the motion parameters.
[0091] Among them, during the forward driving of the vehicle, when the first detection device installed on the vehicle detects the existence of an obstacle in the target area, the first detection device, that is, the monocular camera, will continue to track the obstacle in the target area, obtain the motion parameters of the obstacle, and send these motion parameters to the vehicle's automatic driving system so that the automatic driving system can analyze them and obtain the estimated motion trajectory of the obstacle in the target area.
[0092] S103: Determine whether the vehicle has a collision risk based on the estimated driving trajectory and the estimated motion trajectory; if so, execute step S104; if not, execute step S105.
[0093] Among them, the vehicle's automatic driving system analyzes and processes the vehicle's estimated driving trajectory and the estimated motion trajectory of obstacles in the target area, that is, the vehicle's estimated driving trajectory is compared with the estimated motion trajectory of the obstacle to determine whether there is an overlapping area, so as to judge whether there is a risk of collision between the vehicle and the obstacle.
[0094] S104: Determine whether the obstacle is a target obstacle.
[0095] The target obstacle may be, for example, an obstacle with the highest collision risk value among obstacles that have a collision risk with the vehicle.
[0096] When there is a risk of collision between the vehicle and an obstacle, if there are multiple obstacles that have a risk of collision with the vehicle, it is also necessary to determine whether the obstacle is the target obstacle, that is, the obstacle with the highest collision risk value among the obstacles that have a risk of collision with the vehicle. If the obstacle is the target obstacle, then after the self-check is completed, the vehicle's automatic driving system will finally output the result as the obstacle; if the obstacle is not the target obstacle, it is also necessary to determine whether other obstacles that have a risk of collision with the vehicle are target obstacles, until the target obstacle is determined and an accurate result is output.
[0097] S105: The target obstacle does not exist in the target area.
[0098] Among them, when there is no risk of collision between the vehicle and an obstacle, that is, when the vehicle is moving forward, there is no obstacle that poses a risk of collision with the vehicle within a longitudinal distance of more than 60 meters and a field of view angle in front of the vehicle that is less than 45°. Therefore, there is no target obstacle in the target area and the autonomous driving system cannot output the target obstacle.
[0099] An obstacle identification method provided by an embodiment of the present application, in which a first detection device determines that there is an obstacle in a target area, and the vehicle's automatic driving system obtains the vehicle's driving data to obtain the vehicle's estimated driving trajectory, and continuously tracks the obstacle through the first detection device to obtain the obstacle's motion parameters, and determines the obstacle's estimated motion trajectory based on the motion parameters, and determines whether there is a collision risk between the vehicle and the obstacle based on the vehicle's estimated driving trajectory and the obstacle's estimated motion trajectory. If there is, it is determined whether the obstacle is the target obstacle with the highest collision risk value, until the vehicle's automatic driving system outputs the target obstacle. This method realizes obstacle identification during the vehicle's forward driving process, and enhances the reliability and accuracy of target obstacle detection in complex environments during vehicle driving.
[0100] Figure 3 A process of an obstacle identification method provided in an embodiment of the present application Figure 2 This embodiment is in Figure 2 Based on the embodiment, the obstacle recognition method is described in detail. Figure 3 As shown, this embodiment provides an obstacle recognition method, including:
[0101] S201: When it is determined by the first detection device that there is an obstacle in the target area, obtaining driving data of the vehicle, the driving data including: driving speed and yaw angular velocity.
[0102] Among them, during the forward driving of the vehicle, when the first detection device installed on the inside of the front windshield of the vehicle detects the existence of an obstacle in the target area, that is, when the monocular camera installed on the vehicle detects that there is an obstacle outside the longitudinal distance of more than 60 meters of the vehicle and within the range of the vehicle's field of view angle less than 45° when the vehicle is driving forward, the vehicle's automatic driving system obtains the vehicle's driving data at this time, and the driving data includes the vehicle's own driving speed and yaw angular velocity.
[0103] Optionally, the obstacle recognition method in this embodiment is only used when the vehicle is traveling forward. When the vehicle is in reverse gear or making a U-turn, the obstacle recognition function is in standby mode.
[0104] S202: Determine a driving radius of the vehicle according to the driving speed and the yaw angular velocity.
[0105] The corresponding vehicle speed and yaw rate can be obtained from the vehicle driving data. The quotient of the vehicle speed and yaw rate is the vehicle driving radius. The specific formula is as follows:
[0106]
[0107] Where R is the vehicle's driving radius, v is the vehicle's driving speed, and ω is the vehicle's yaw angular velocity.
[0108] Optionally, after obtaining the driving radius based on the vehicle's driving data, it is necessary to confirm whether the driving radius is greater than 150 meters. If the vehicle's driving radius is greater than 150 meters, the vehicle is in a normal forward driving state, and the vehicle's obstacle recognition function is activated; if the vehicle's driving radius is less than 150 meters, the vehicle is not in a normal forward driving state or is in reverse gear and U-turn state, and the vehicle's obstacle recognition function remains on standby until the vehicle enters the normal forward driving process.
[0109] S203: Determine an estimated driving trajectory of the vehicle according to the driving radius.
[0110] Among them, after the vehicle's automatic driving system calculates the vehicle's driving radius, it can estimate the estimated driving trajectory of the vehicle corresponding to it based on the radius. The purpose of determining the estimated driving trajectory of the vehicle in this step is to determine whether there is a risk of collision between the vehicle and the obstacle based on the estimated driving trajectory of the vehicle and the estimated motion trajectory of the obstacle in the subsequent steps.
[0111] It is understandable that in actual applications, the vehicle status and driving environment are constantly changing. Therefore, the estimated driving trajectory needs to be updated in real time to reflect the latest driving status and environmental information.
[0112] S204: Acquire motion parameters of the obstacle through the first detection device, where there are multiple obstacles, and the motion parameters include: position, motion speed, and orientation angle.
[0113] Among them, when the vehicle's obstacle recognition function is activated and there is an obstacle in front of the vehicle with a field of view angle less than 45° and a longitudinal distance greater than 60m from the vehicle, the vehicle's automatic driving system continuously tracks the obstacle's position, movement speed and orientation angle information through the first detection device, that is, a monocular camera installed on the inside of the vehicle's front windshield. Since there may be more than one obstacle and each obstacle acquired by the monocular camera has motion parameters, the vehicle's automatic driving system stores the position, movement speed and orientation angle information of multiple obstacles in the target area.
[0114] S205: For any one obstacle among the multiple obstacles, determine an estimated motion trajectory of the obstacle according to the position, motion speed, and orientation angle.
[0115] Among them, for any one of the multiple obstacles, the vehicle's automatic driving system can estimate the estimated motion trajectory of the obstacle based on the obstacle position, movement speed and orientation angle information output by the monocular camera. The purpose of obtaining the estimated motion trajectory of the obstacle in this step is to compare it with the estimated driving trajectory of the vehicle to determine whether there is a risk of collision between the vehicle and the obstacle.
[0116] S206: Determine the motion state of each obstacle according to the estimated motion trajectories of the multiple obstacles, where the motion state includes: traversing motion, same-direction motion, and stationary motion.
[0117] Among them, if the estimated motion trajectory of the obstacle intersects with the estimated driving trajectory of the vehicle, the motion state of the obstacle is lateral motion; if the estimated motion trajectory of the obstacle is parallel to the estimated driving trajectory of the vehicle, the motion state of the obstacle is co-directional motion; if the estimated motion trajectory of the obstacle is a point, the obstacle is in a stationary state.
[0118] Based on the estimated motion trajectories of multiple obstacles, the vehicle's autonomous driving system classifies the motion state of each obstacle, including: lateral motion, same-direction motion, and stationary.
[0119] It is understandable that the movement of obstacles may be affected by the surrounding environment, such as road shape, traffic signals, the presence of other obstacles, etc. The impact of these external factors on the movement trajectory needs to be considered during prediction. Therefore, the movement state of obstacles is classified to more accurately predict the future movement trajectory of obstacles, so as to effectively avoid collisions and plan safe driving routes.
[0120] S207: For an obstacle whose motion state is lateral motion or same-direction motion, determine a collision risk area of the obstacle according to the estimated motion trajectory of the obstacle and the estimated driving trajectory.
[0121] Among them, when the movement state of the obstacle is lateral movement or co-directional movement, the possible driving path of the vehicle within a certain period of time can be predicted based on the estimated driving trajectory of the vehicle. Based on the estimated movement trajectory of the obstacle, the possible position of the obstacle within the same time range can be predicted. According to the possible driving path of the vehicle and the possible position of the obstacle within the specified time period, the collision risk area between the obstacle and the vehicle can be determined, that is, the area where the possible driving path of the vehicle overlaps with the possible position of the obstacle.
[0122] It is understood that within the selected time period, it is determined whether there is an overlapping area between the estimated motion trajectory of the obstacle and the estimated driving trajectory of the vehicle. If at any time within the time period, the minimum safe distance between the two is less than a preset threshold, then the area is considered as a potential collision risk area.
[0123] S208: For an obstacle whose motion state is stationary, the position of the obstacle is used as a collision risk area of the obstacle.
[0124] Among them, when the movement state of the obstacle is stationary, that is, when the obstacle is stationary, determine whether the position of the obstacle within a certain period of time coincides with the estimated driving trajectory of the vehicle within the period of time. If there is an overlap, the area where the obstacle is located is used as the collision risk area of the obstacle within the specified period of time.
[0125] It can be understood that if the position of the obstacle within a certain period of time does not coincide with the estimated driving trajectory of the vehicle within the period of time, the monocular camera needs to continuously track the obstacle and judge whether the position of the obstacle in the corresponding period of time coincides with the estimated driving trajectory of the vehicle within the period of time according to the movement state of the obstacle in different time periods, so as to determine whether there is a collision risk between the obstacle and the vehicle, until the obstacle leaves the target area, that is, is located in front of the vehicle with a field of view angle less than 45° and a longitudinal distance greater than 60m from the vehicle.
[0126] S209: Determine whether there is a target collision risk area in the estimated driving trajectory among the multiple collision risk areas; if so, execute step S210; if not, execute step S211.
[0127] Among them, in different time periods, the collision risk area of the obstacle changes with the movement state of the obstacle in the current time period. Therefore, the vehicle's automatic driving system needs to perform real-time detection of the collision risk area of the obstacle, and determine whether the collision risk area of the obstacle overlaps with the estimated driving trajectory of the vehicle. The area overlapping with the estimated driving trajectory of the vehicle is the target collision risk area, and the collision risk area is located within the target collision risk area.
[0128] S210: Determine that the vehicle has a collision risk.
[0129] Among them, when the collision risk area of the obstacle is in the target collision risk area of the estimated driving trajectory, that is, the collision risk area of the obstacle overlaps with the estimated driving trajectory of the vehicle, it means that there is a collision risk between the obstacle and the vehicle and the overlapping area is the target collision risk area.
[0130] It can be understood that since there are multiple obstacles, when the collision risk area of an obstacle exists in the target collision risk area, the vehicle has a collision risk with the obstacle, that is, the vehicle has a collision risk.
[0131] S211: There is no collision risk for the vehicle.
[0132] Among them, when the collision risk areas of all obstacles detected by the monocular camera are not in the target collision risk area of the estimated driving trajectory, there is no collision risk between the vehicle and all detected obstacles, that is, there is no collision risk for the vehicle.
[0133] S212: Determine a plurality of obstacles corresponding to a plurality of target collision risk areas, and determine a collision moment of each obstacle according to the estimated running trajectories of the plurality of obstacles and the estimated driving trajectory of the vehicle.
[0134] Among them, when the vehicle is at risk of collision, multiple obstacles that are at risk of collision with the vehicle are determined, and the vehicle's automatic driving system can estimate the collision time of each obstacle based on the possible position of each obstacle within a certain period of time and the possible driving path of the vehicle within the same period of time.
[0135] S213: Sort the obstacles by the plurality of collision moments in ascending order.
[0136] Among them, the obstacles that have a collision risk with the vehicle are sorted in ascending order of collision time. The collision time of the obstacles in the front order is closer to the current time, that is, the obstacles in the front order will hit the vehicle earlier than the obstacles in the back order. Therefore, the collision risk value of the obstacles in the front order is higher than the collision risk value of the obstacles in the back order.
[0137] S214: taking obstacles whose serial numbers meet preset conditions as candidate obstacles.
[0138] The preset conditions include: the sequence number is less than or equal to the preset sequence number. The preset sequence number may be 1 or 2, for example, that is, the candidate obstacle is the obstacle ranked first or the obstacle ranked first and second. When the candidate obstacle is the obstacle ranked first, the collision time of the obstacle is the smallest and will collide with the vehicle first; when the candidate obstacle is the obstacle ranked first and second, the collision time of the obstacles is the smallest two, that is, the two obstacles that will collide with the vehicle first.
[0139] It can be understood that if the obstacle with the preset sequence number 1 is detected to have an error by the millimeter wave radar, the preset sequence number is shifted down, and the candidate obstacle is the obstacle with the sequence number 2.
[0140] S215: Detecting the real-time movement trajectory of the candidate obstacle in real time through a second detection device.
[0141] The second detection device refers to a millimeter wave radar installed near the first detection device.
[0142] After the vehicle's autonomous driving system screens out candidate obstacles, it needs to use millimeter-wave radar to detect the real-time motion trajectory of the obstacle to prevent misidentification by the monocular camera.
[0143] It is understandable that in the process of continuous tracking of candidate obstacles by a monocular camera, in order to avoid target loss or misidentification due to occlusion, reflection and other problems, millimeter-wave radar is needed for radar target matching.
[0144] S216: If the real-time motion trajectory of the candidate obstacle coincides with the estimated motion trajectory within a preset time period, the candidate obstacle is determined to be the target obstacle.
[0145] Among them, if the real-time motion trajectory of the candidate obstacle coincides with the estimated motion trajectory within the preset time, that is, the real-time motion trajectory of the candidate obstacle has corresponding radar points, then the candidate obstacle is correct, and the vehicle's automatic driving system finally outputs the candidate obstacle as the target obstacle.
[0146] Optionally, in any case, if the driver manually turns off the camera, the obstacle recognition function will be turned off.
[0147] The embodiment of the present application provides an obstacle recognition method, which obtains the motion parameters of obstacles in the target area detected by a monocular camera, determines the estimated motion trajectory of all obstacles, and then determines the motion state according to the different estimated motion trajectories of the obstacles. The collision risk area of each obstacle can be obtained from the real-time motion state of each obstacle and the real-time estimated driving trajectory of the vehicle. If the collision risk area of an obstacle coincides with the estimated driving trajectory of the vehicle during the real-time movement of the vehicle, the vehicle has a collision risk, and the overlapping area is the target collision risk area. All obstacles in the target collision risk area are sorted, and the obstacle that the vehicle will collide with first is used as a candidate obstacle. The candidate obstacle is then detected and processed by a millimeter-wave radar. After eliminating the possibility of misidentification, the vehicle's automatic driving system outputs the target obstacle, thereby achieving the recognition of the obstacle. This method uses a monocular camera and a forward millimeter-wave radar for fusion perception. Compared with automatic driving vehicles whose target information is mostly derived from a single sensor, it is easier to realize the system's recognition, tracking and collision risk analysis of target obstacles in front of the vehicle in complex traffic scenes.
[0148] Figure 4 This is a schematic diagram of the structure of an obstacle recognition device provided in this application. Figure 4 As shown, this embodiment provides an obstacle identification device 400, which includes:
[0149] The acquisition module 401 is used to acquire the driving data of the vehicle when it is determined by the first detection device that there is an obstacle in the target area;
[0150] A determination module 402, configured to determine an estimated driving trajectory of the vehicle based on the driving data;
[0151] The acquisition module 401 is further configured to acquire the motion parameters of the obstacle through the first detection device;
[0152] The determination module 402 is further configured to determine an estimated motion trajectory of the obstacle based on the motion parameters;
[0153] A judgment module 403, judging whether the vehicle has a collision risk according to the estimated driving trajectory and the estimated motion trajectory;
[0154] The determination module 402 is further configured to determine whether the obstacle is a target obstacle when there is a collision risk with the vehicle.
[0155] Optionally, the driving data includes: driving speed and yaw angular velocity;
[0156] The determination module 402 is further configured to determine a driving radius of the vehicle according to the driving speed and the yaw angular velocity; and determine an estimated driving trajectory of the vehicle according to the driving radius.
[0157] Optionally, there are multiple obstacles, and the motion parameters include: position, motion speed and orientation angle;
[0158] The determination module 402 is further configured to determine, for any one of the multiple obstacles, an estimated motion trajectory of the obstacle according to the position, motion speed, and orientation angle.
[0159] Optionally, the determination module 402 is further configured to determine the motion state of each obstacle according to the estimated motion trajectories of the multiple obstacles, wherein the motion state includes: traversing motion, same-direction motion, and stationary motion; and determine the collision risk area of each obstacle according to the motion states of the multiple obstacles;
[0160] The judgment module 403 is further used to judge whether there is a target collision risk area in the estimated driving trajectory in the multiple collision risk areas;
[0161] The determination module 402 is further configured to determine that the vehicle has a collision risk when there is a target collision risk area located in the estimated driving trajectory among the multiple collision risk areas.
[0162] Optionally, the device further includes a processing module 404;
[0163] The determination module 402 is further configured to determine, for an obstacle in a transverse motion or in a same direction motion state, a collision risk area of the obstacle according to the estimated motion trajectory of the obstacle and the estimated driving trajectory;
[0164] The processing module 404 is configured to use the position of an obstacle that is in a stationary state as a collision risk area of the obstacle.
[0165] Optionally, the determination module 402 is further configured to determine a candidate obstacle from obstacles corresponding to a plurality of target collision risk areas;
[0166] The processing module 404 is further configured to detect the real-time motion trajectory of the candidate obstacle in real time through a second detection device;
[0167] The determination module 402 is further configured to determine that the candidate obstacle is the target obstacle if the real-time motion trajectory of the candidate obstacle coincides with the estimated motion trajectory within a preset time period.
[0168] Optionally, the determination module 402 is further configured to determine a plurality of obstacles corresponding to a plurality of target collision risk areas, and determine a collision moment of each obstacle according to the estimated running trajectories of the plurality of obstacles and the estimated driving trajectory of the vehicle;
[0169] The processing module 404 is further configured to sort the obstacles according to the order of the multiple collision moments from small to large, wherein the collision risk value of the obstacle in the front arrangement order is higher than the collision risk value of the obstacle in the back arrangement order; and to select the obstacle whose serial number meets the preset condition as the candidate obstacle, wherein the preset condition includes: the serial number is less than or equal to the preset serial number.
[0170] Figure 5 This is a schematic diagram of the structure of an obstacle recognition device provided in this application. Figure 5 As shown, the present application provides an obstacle recognition device, and the obstacle recognition device 500 includes: a receiver 501, a transmitter 502, a processor 503 and a memory 504.
[0171] Receiver 501, used for receiving instructions and data;
[0172] A transmitter 502, used for sending instructions and data;
[0173] Memory 504, used to store computer-executable instructions;
[0174] The processor 503 is used to execute the computer-executable instructions stored in the memory 504 to implement the various steps performed by the obstacle recognition method in the above embodiment. For details, please refer to the relevant description in the above obstacle recognition method embodiment.
[0175] Optionally, the memory 504 may be independent or integrated with the processor 503 .
[0176] When the memory 504 is independently provided, the electronic device further includes a bus for connecting the memory 504 and the processor 503 .
[0177] The present application also provides a computer storage medium, in which computer execution instructions are stored. When a processor executes the computer execution instructions, the obstacle recognition method executed by the above-mentioned obstacle recognition device is implemented.
[0178] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical nodes; for example, a physical node may have multiple functions, or a function or step may be performed by several physical nodes in cooperation. Some physical nodes or all physical nodes may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.
[0179] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.
[0180] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.
Claims
1. An obstacle recognition method, It is characterized in that include: When it is determined by the first detection device that there is an obstacle in the target area, obtaining driving data of the vehicle, and determining an estimated driving trajectory of the vehicle based on the driving data; Acquire motion parameters of the obstacle through the first detection device, and determine an estimated motion trajectory of the obstacle based on the motion parameters; Determining whether the vehicle has a collision risk according to the estimated driving trajectory and the estimated motion trajectory; If there is a collision risk with the vehicle, it is determined whether the obstacle is a target obstacle.
2. The method according to claim 1, It is characterized in that The driving data includes: driving speed and yaw angular velocity, and determining the estimated driving trajectory of the vehicle based on the driving data includes: Determining a driving radius of the vehicle according to the driving speed and the yaw angular velocity; An estimated driving trajectory of the vehicle is determined according to the driving radius.
3. The method according to claim 1, It is characterized in that There are multiple obstacles, and the motion parameters include: position, motion speed and orientation angle. The step of determining the estimated motion trajectory of the obstacle based on the motion parameters includes: For any one obstacle among the multiple obstacles, an estimated motion trajectory of the obstacle is determined according to the position, motion speed and orientation angle.
4. The method according to claim 3, It is characterized in that The determining, according to the estimated driving trajectory and the estimated motion trajectory, whether the vehicle has a collision risk includes: Determine the motion state of each obstacle according to the estimated motion trajectories of the multiple obstacles, wherein the motion state includes: traversing motion, same-direction motion, and stationary motion; According to the motion states of multiple obstacles, the collision risk area of each obstacle is determined; determining whether there is a target collision risk area in the estimated driving trajectory among the multiple collision risk areas; If so, it is determined that the vehicle has a collision risk.
5. The method according to claim 4, It is characterized in that The step of determining the collision risk area of each obstacle according to the motion states of the plurality of obstacles includes: For an obstacle whose motion state is lateral motion or same-direction motion, determining a collision risk area of the obstacle according to the estimated motion trajectory of the obstacle and the estimated driving trajectory; For an obstacle whose motion state is stationary, the position of the obstacle is used as the collision risk area of the obstacle.
6. The method according to claim 1, It is characterized in that If the vehicle has a collision risk, determining whether the obstacle is a target obstacle includes: Determine a candidate obstacle from obstacles corresponding to a plurality of target collision risk areas; Detecting the real-time motion trajectory of the candidate obstacle in real time by a second detection device; If the real-time motion trajectory of the candidate obstacle coincides with the estimated motion trajectory within a preset time period, the candidate obstacle is determined to be the target obstacle.
7. The method according to claim 6, It is characterized in that The determining of candidate obstacles from obstacles corresponding to a plurality of target collision risk areas includes: Determine a plurality of obstacles corresponding to a plurality of target collision risk areas, and determine a collision moment of each obstacle according to the estimated running trajectories of the plurality of obstacles and the estimated driving trajectory of the vehicle; sorting the obstacles in ascending order of the multiple collision moments, wherein the collision risk value of the obstacle in the front order is higher than the collision risk value of the obstacle in the back order; The obstacles whose serial numbers meet a preset condition are taken as the candidate obstacles, wherein the preset condition includes: the serial number is less than or equal to a preset serial number.
8. An obstacle recognition device, It is characterized in that The device comprises: An acquisition module, used to acquire the driving data of the vehicle when it is determined by the first detection device that there is an obstacle in the target area; A determination module, configured to determine an estimated driving trajectory of the vehicle based on the driving data; The acquisition module is further used to acquire the motion parameters of the obstacle through the first detection device; The determination module is further used to determine the estimated motion trajectory of the obstacle based on the motion parameters; A judgment module, for judging whether the vehicle has a collision risk according to the estimated driving trajectory and the estimated motion trajectory; The determination module is further configured to determine whether the obstacle is a target obstacle when there is a collision risk with the vehicle.
9. An obstacle recognition device, It is characterized in that include: Memory; processor; Wherein, the memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the obstacle recognition method according to any one of claims 1 to 7.
10. A computer storage medium, It is characterized in that The computer storage medium stores computer executable instructions, and when the computer executable instructions are executed by the processor, they are used to implement the obstacle recognition method according to any one of claims 1 to 7.
