Straight way cut-in target identification method and device, electronic equipment and medium

Abstract

The invention provides a straightway cut-in target identification method and device, electronic equipment and a medium, and the method comprises the steps: determining a target obstacle from all obstacles according to the relative position relation between a vehicle and all obstacles; determining a plurality of key point coordinates of the target obstacle according to the relative position coordinates, the motion information and the appearance information of the target obstacle; a target key point closest to the vehicle in the transverse direction is determined from the multiple key point coordinates of the target obstacle; determining a main target from the target obstacles according to the determined target key points of the target obstacles; and controlling the vehicle to cruise based on the main target. The objective of the invention is to improve the opportunity accuracy of selecting a main target followed by a vehicle so as to reduce the driving risk.

Description

Technical Field

[0001] The present application relates to the field of assisted driving technology, and in particular to a method, device, electronic device and medium for identifying a straight-line cut-in target. Background Art

[0002] Currently, there are two main modes of ACC (Adaptive Cruise Control). One is cruise control, which allows the assisted driving vehicle to drive automatically by setting a fixed speed. The other is to select the ACC main target in the lane (that is, select a car in front of the vehicle as the target to follow), follow the main target, and maintain a corresponding safe distance to achieve automatic driving of the vehicle. The second method is the main cruising mode of the current ACC system.

[0003] For the second cruising mode, when an obstacle vehicle cuts into the lane where the ego vehicle is located from the adjacent lane of the lane where the ego vehicle is located, and the cut-in position is between the ego vehicle and the main target currently followed by the ego vehicle, the obstacle vehicle will be selected as the new main target to be followed by the ego vehicle. In this process, if the vehicle is identified by the traditional target recognition method used in the second cruising mode (that is, the sensed vehicle is identified as a point), the timing of selecting the obstacle vehicle as the new main target to be followed by the ego vehicle will be late, which will increase the probability of collision between the ego vehicle and the obstacle vehicle, thereby increasing the driving risk. For example, the midpoint of the rear of the obstacle vehicle is identified as the obstacle vehicle. When the obstacle vehicle intervenes in the lane where the ego vehicle is located, the front of the obstacle vehicle will enter the lane where the ego vehicle is located before the midpoint of the rear of the obstacle vehicle, which will cause the obstacle vehicle to be selected as the new main target to be followed by the ego vehicle to be late. Summary of the invention

[0004] In view of this, the present application provides a straight-line cut-in target recognition method, device, electronic device and medium, aiming to improve the accuracy of the timing of selecting the main target followed by the vehicle, so as to reduce driving risks.

[0005] The present application provides a straight cut-in target recognition method, the method comprising:

[0006] Determining a target obstacle from the obstacles according to the relative positional relationship between the vehicle and the obstacles;

[0007] Determine the coordinates of multiple key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle;

[0008] Determine the target key point that is closest to the vehicle in the lateral direction from the coordinates of the multiple key points of the target obstacle;

[0009] Determining a main target from among the target obstacles according to the determined target key points of the target obstacles;

[0010] The vehicle is controlled to cruise based on the main target.

[0011] Optionally, determining the main target from each target obstacle according to the determined target key points of each target obstacle includes:

[0012] Determine the lateral distance between the target key points of each target obstacle and the ego vehicle according to the determined target key points of each target obstacle;

[0013] Determine the relationship between the lateral distance between each target key point of each target obstacle and the ego vehicle and the corresponding preset distance range;

[0014] When there is a target key point whose lateral distance to the vehicle is within the preset distance range corresponding to the vehicle, the target obstacle corresponding to the target key point is determined as the main target.

[0015] Optionally, the method further includes:

[0016] According to the relative speed between the target obstacle and the vehicle, a preset distance range corresponding to the target obstacle is determined.

[0017] Optionally, when there are multiple target key points and the lateral distances between the vehicle and the target key points are within the corresponding preset distance ranges, the method further includes:

[0018] Determine the target obstacles corresponding to each of the plurality of target key points as candidate main targets;

[0019] Determine the longitudinal distance between each candidate main target and the ego vehicle;

[0020] The candidate main target corresponding to the smallest longitudinal distance is determined as the main target.

[0021] Optionally, determining the longitudinal distance between each candidate main target and the vehicle includes:

[0022] Determine the first target key point that is closest to the vehicle in the longitudinal direction from the coordinates of the multiple key points of each candidate main target;

[0023] The longitudinal distance between the first target key point and the vehicle is determined as the longitudinal distance between the candidate main target corresponding to the first target key point and the vehicle.

[0024] Optionally, the method further includes:

[0025] When the length in the shape information of the target obstacle exceeds a first threshold, setting the length in the shape information of the target obstacle to a preset length;

[0026] When the width in the shape information of the target obstacle exceeds a second threshold, the width in the shape information of the target obstacle is set to a preset width.

[0027] Optionally, before determining the coordinates of a plurality of key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle, the method further includes:

[0028] According to the motion information of the target obstacle, determine whether the target obstacle has the intention to cut into the lane where the ego vehicle is located;

[0029] Target obstacles that do not intend to cut into the lane where the vehicle is located are deleted from the target obstacle queue.

[0030] Compared with the prior art, this application has the following advantages:

[0031] The present application provides a straight-line target recognition method, which determines the target obstacle from each obstacle according to the relative position relationship between the self-vehicle and each obstacle; determines the coordinates of multiple key points of the target obstacle according to the relative position coordinates, motion information and appearance information of the target obstacle; determines the target key point that is closest to the self-vehicle in the lateral direction from the multiple key point coordinates of the target obstacle; determines the main target from each target obstacle according to the determined target key points of each target obstacle; and controls the self-vehicle to cruise based on the main target. Thus, by identifying the obstacle as a frame target that is closer to the obstacle body, the target key point that is closest to the lateral direction of the frame target is calculated based on the Bounding Box, and then the main target is determined from all target obstacles based on the target key point of the target obstacle. Since the frame target is closer to the obstacle body, the main target that the self-vehicle will follow can be selected at a more accurate time to reduce driving risks. At the same time, compared with the traditional target recognition method, the present invention is simpler to implement, consumes less computing power, and saves memory usage.

[0032] The second aspect of the present application provides a straight-line cut-in target recognition device, which aims to improve the accuracy of the timing of selecting the main target followed by the vehicle, so as to reduce driving risks.

[0033] A second aspect of the present application provides a straight-line cut-in target recognition device, the device comprising:

[0034] A target obstacle determination unit, configured to determine a target obstacle from among the obstacles according to a relative position relationship between the vehicle and each obstacle;

[0035] A key point coordinate determination unit, used to determine the coordinates of multiple key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle;

[0036] A target key point determination unit, used to determine the target key point that is closest to the vehicle in the lateral direction from the coordinates of the multiple key points of the target obstacle;

[0037] A main target determination unit, configured to determine a main target from each target obstacle according to the determined target key points of each target obstacle;

[0038] A control unit is used to control the vehicle to cruise based on the main target.

[0039] A third aspect of the present application provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0040] Memory, used to store computer programs;

[0041] The processor is used to implement the steps of the straight-line target identification method described in the first aspect of the present invention when executing the program stored in the memory.

[0042] Optionally, the fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the straight-line target identification method described in the first aspect of the present application are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0044] Figure 1 is a flow chart of a straight-line cut-in target identification method provided in an embodiment of the present application;

[0045] Figure 2 is a schematic diagram of a target reference point in a straight-line cut-in target recognition method shown in an embodiment of the present application;

[0046] Figure 3 is a schematic diagram of key points in a straight-line cut-in target recognition method shown in an embodiment of the present application;

[0047] Figure 4is another flow chart of a straight-line cut-in target identification method provided by an embodiment of the present application;

[0048] Figure 5 It is a schematic diagram of a straight-line cut-in target recognition device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] The exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0050] Figure 1 is a flow chart of a straight-line target identification method provided by an embodiment of the present application, such as Figure 1 As shown, the method includes:

[0051] Step S101: determining a target obstacle from the obstacles according to the relative positional relationship between the vehicle and the obstacles;

[0052] Step S102: determining the coordinates of multiple key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle;

[0053] Step S103: determining the target key point that is closest to the vehicle in the lateral direction from the coordinates of the multiple key points of the target obstacle;

[0054] Step S104: determining a main target from each target obstacle according to the determined target key points of each target obstacle;

[0055] Step S105: Control the vehicle to cruise based on the main target.

[0056] In the embodiment of the present invention, since there are many obstacles around the vehicle during the actual driving process, if the main target is determined from all the obstacles, a large amount of computing resources will be required for calculation, resulting in a serious waste of computing resources. Therefore, the present invention first selects a part of the target obstacles closest to the vehicle from the many obstacles around the vehicle, and determines the main target among the target obstacles after the range is narrowed. The present invention first determines the target obstacles that will affect the driving of the vehicle from the target obstacles according to the relative position relationship between the vehicle and each obstacle that the vehicle can monitor.

[0057] In an embodiment of the present invention, an optional implementation of step S101 is: since the millimeter wave radar has low accuracy in identifying the relative lateral distance of the target, while the camera can accurately identify the lateral distance, length, width and heading angle of the target, in order to accurately identify the target in front of the vehicle, the present invention collects obstacle information through the front millimeter wave radar and the front camera of the vehicle. Based on the obstacle information identified by the front millimeter wave radar and the front camera, the center of the rear bumper of the obstacle is determined as the target reference point of the obstacle, such as Figure 2 As shown in FIG. 1 , obstacles include cars and trucks traveling on the motor vehicle lane.

[0058] In an embodiment of the present invention, after determining the target reference point of each obstacle through obstacle information collected by the front millimeter-wave radar and the front camera of the self-vehicle, a self-vehicle coordinate system is established with the center of the rear axle of the self-vehicle as the origin, and the relative position coordinates of the target reference point of each obstacle to the origin of the self-vehicle coordinate system are calculated in the self-vehicle coordinate system, that is, the relative longitudinal distance dx and the relative lateral distance dy from the target reference point of the obstacle to the center of the rear axle of the self-vehicle, thereby obtaining the relative position coordinates of each obstacle around the self-vehicle body and the self-vehicle through calculation.

[0059] According to the calculated relative position coordinates between each obstacle and the vehicle, the obstacle closest to the vehicle in front of the lane where the vehicle is located is determined as the target obstacle, the two obstacles closest to the vehicle in front of the vehicle in the left lane adjacent to the vehicle's lane are determined as target obstacles, and the two obstacles closest to the vehicle in front of the vehicle in the right lane adjacent to the vehicle's lane are determined as target obstacles. For example, there are 4 obstacles in front of the vehicle in the left lane adjacent to the vehicle's lane, obstacle No. 1 is 3m away from the vehicle, obstacle No. 2 is 4m away from the vehicle, obstacle No. 3 is 2m away from the vehicle, and obstacle No. 4 is 6m away from the vehicle, so No. 1 and No. 3 are selected as target obstacles. Thus, according to the relative position coordinates between each obstacle and the vehicle, the target obstacle is determined from each obstacle.

[0060] In an embodiment of the present invention, in order to more accurately determine a switching time for the self-vehicle to switch to the main target of the cruise and reduce driving risks, the present invention determines the target obstacle as a frame target with length and width. Since the frame target is closer to the main structure of the target obstacle, the main target is selected based on the frame target of the target obstacle, and a more accurate switching time can be determined for the self-vehicle to switch to the main target of the cruise. Specifically, based on the relative position coordinates of the target obstacle, the heading angle in the motion information, and the length and width shape information of the target obstacle determined by the front millimeter-wave radar and the front camera, multiple key point coordinates of the target obstacle are determined, wherein the multiple key point coordinates refer to the coordinates of the four vertices of the target obstacle in the self-vehicle coordinate system, such as Figure 3As shown in the figure, LL, LU, RU, and RL are the four vertices of the target obstacle, that is, the four key points. The coordinates corresponding to these four key points in the vehicle coordinate system are the key point coordinates. For each target obstacle, there are four key points. Among them, the body shape of the vehicle is symmetrical. When only three key points of the target obstacle are obtained, the key point coordinates of the fourth key point of the target obstacle can be calculated based on the key point coordinates of the three key points.

[0061] In the embodiment of the present invention, the four key points are calculated by a Bounding Box (Bounding Box regression technology, used to locate the bounding box of the target in the image), and the Bounding Box is composed of coordinates containing four parameters, namely G = (G x ,G y ,G w ,G h ), where G x , G y Indicates the horizontal and vertical coordinates of the center point of the Bounding Box, G w , G h Respectively represent the width and height of the Bounding Box.

[0062] In the embodiment of the present invention, after obtaining the coordinates of the four key points of each target obstacle, for the target obstacle, a key point that is closest to the lateral distance of the vehicle is determined from the coordinates of the four key points of the target obstacle as the target key point of the target obstacle, which is used for the subsequent determination of the main target. For each target obstacle, its own target key point is determined in this implementation. Thus, one target obstacle has one target key point.

[0063] In an embodiment of the present invention, after obtaining a target key point corresponding to each target obstacle, a main target is determined from all the determined target obstacles according to the determined target key points.

[0064] In the embodiment of the present invention, after the main target is determined, the vehicle is controlled to perform automatic cruising based on the main target according to the motion state of the main target. For example, when the main target is moving normally, the vehicle follows the main target; when the main target is too close to the vehicle or emergency braking occurs, the vehicle brakes or switches lanes to avoid danger.

[0065] In the embodiments of the present invention, the lateral distance of obstacles on the curve cannot reflect the relationship between the obstacles and the lane line. Relying solely on the lateral distance will frequently misselect obstacles in the adjacent lane that are driving normally, resulting in frequent braking of the vehicle and poor driving experience. Therefore, the present invention is aimed at the selection of obstacles that cut into the straight road, that is, the applicable scenario of the present invention is the straight road. Whether the current road on which the vehicle is traveling is a straight road can be judged by the curvature of the road, and the straight road scene can be limited by limiting the curvature of the road.

[0066] The present application provides a straight-line target recognition method, which collects the target reference point of the obstacle through the millimeter-wave radar and front camera of the vehicle. Then, the vehicle coordinate system is established with the rear axle center of the vehicle as the origin, and the relative position coordinates of the target reference points of the obstacles to the origin are calculated; the target obstacle is determined from the obstacles according to the collected lane line information and the relative position coordinates of the obstacles; the multiple key point coordinates of the target obstacle are determined according to the relative position coordinates, motion information and shape information of the target obstacle; the target key point with the closest lateral distance to the vehicle is determined from the multiple key point coordinates of the target obstacle; the main target is determined from the target obstacles according to the determined target key points of each target obstacle; and the vehicle is controlled to cruise based on the main target. Thus, by identifying the obstacle as a frame target that is closer to the obstacle body, the target key point with the closest lateral distance to the frame target is calculated based on the Bounding Box, and then the main target is determined from all target obstacles based on the target key points of the target obstacle. Since the frame target is closer to the obstacle body, the accuracy of selecting the main target followed by the vehicle can be improved to reduce driving risks. At the same time, compared with traditional target recognition methods, the present invention is simpler to implement, consumes less computing power, and saves memory usage.

[0067] In the present invention, the method also includes: when the length in the shape information of the target obstacle exceeds a first threshold, setting the length in the shape information of the target obstacle to a preset length; when the width in the shape information of the target obstacle exceeds a second threshold, setting the width in the shape information of the target obstacle to a preset width.

[0068] In the embodiment of the present invention, the present invention finds that for obstacles with large length and width such as trucks, the closest point of the lateral distance of the obstacle obtained is easy to exceed the road width of the lane where the vehicle is located, that is, the closest point of the lateral distance is not in the lane, so that the closest point of the lateral distance of such obstacles identified is far away from the vehicle and has exceeded the lane where the vehicle is located, so that the obstacle that may be selected as the main target cannot be selected. To solve this problem, the present invention limits the length and width of the target obstacle. Specifically, when the length in the shape information of the target obstacle exceeds the first threshold, the length in the shape information of the target obstacle is set to a preset length; when the width in the shape information of the target obstacle exceeds the second threshold, the width in the shape information of the target obstacle is set to a preset width. Preferably, the first threshold is preferably 6m, the second threshold is preferably 4m, the preset length is preferably 6m, and the preset width is preferably 4m. It should be understood that the above-mentioned first threshold, second threshold, preset length and preset width are only preferred values ​​and are not limited to the present invention. The above-mentioned first threshold, second threshold, preset length and preset width of the present invention can also be set to other values ​​according to the actual application scenario, and are not specifically limited here.

[0069] In the present invention, the determining of the main target from each target obstacle according to the determined target key points of each target obstacle includes: determining the lateral distances between each target key point of each target obstacle and the ego vehicle according to the determined target key points of each target obstacle; determining the relationship between the lateral distances between each target key point of each target obstacle and the ego vehicle and the corresponding preset distance ranges; in the case where there is a target key point whose lateral distance to the ego vehicle is within the corresponding preset distance range, determining the target obstacle corresponding to the target key point as the main target.

[0070] In an embodiment of the present invention, an implementation of the above step S104 is: after determining a target key point of each target obstacle that is closest to the vehicle, calculate the lateral distance between each target key point and the origin of the vehicle. Then set a preset distance range for each target obstacle. Compare the lateral distance between each target key point of each target obstacle and the origin of the vehicle with the corresponding preset distance range. When the lateral distance between the target key point of only one target obstacle and the vehicle is within the preset distance range corresponding to the target obstacle, the target obstacle is determined as the main target. For example, the target obstacles include target obstacle 1, and the corresponding preset distance range is set to [0, 2], target obstacle 2, and the corresponding preset distance range is set to [0, 1.5], target obstacle 3 and the corresponding preset distance range is set to [0, 2.2], target obstacle 4 and the corresponding preset distance range is set to [0, 1.8], and target obstacle 5 and the corresponding preset distance range is set to [0, 2.4]; then, the distance between the target key point of target obstacle 1 and the origin of the vehicle is determined to be 1.2, the distance between the target key point of target obstacle 2 and the origin of the vehicle is determined to be 2, the distance between the target key point of target obstacle 3 and the origin of the vehicle is determined to be 2.5, the distance between the target key point of target obstacle 4 and the origin of the vehicle is determined to be 2, and the distance between the target key point of target obstacle 5 and the origin of the vehicle is determined to be 2.5; thus, it is determined that only the distance between the target obstacle 1 and the origin of the vehicle is within the preset distance range corresponding to itself, and therefore the target obstacle is determined as the main target.

[0071] In the present invention, the method further comprises: determining a preset distance range corresponding to the target obstacle according to a relative speed between the target obstacle and the vehicle.

[0072] In the embodiment of the present invention, the present application finds that when an obstacle approaches the lane where the vehicle is located, the risk caused by the obstacle to the vehicle is not only related to the distance of the obstacle from the vehicle, but also to the relative speed between the obstacle and the vehicle. For example, there are two target obstacles located in front of the vehicle, and the lateral distances from the vehicle to the vehicle are 1m and 1.5m respectively. The target obstacle with a lateral distance of 1m from the vehicle is higher than the speed of the vehicle, while the target obstacle with a lateral distance of 1.5m from the vehicle is lower than the speed of the vehicle. Therefore, the target obstacle with a lateral distance of 1.5m from the vehicle will be more dangerous to the vehicle. Therefore, the target obstacle is set to a larger preset distance range than the target obstacle with a lateral distance of 1m from the vehicle, which will better ensure the safety of the vehicle. Therefore, the present invention first determines the relative speed between the target obstacle in the lane where the vehicle is located and the two adjacent lanes on the left and right of the lane and the vehicle. The preset distance range of the target obstacle with a speed less than the speed of the vehicle is set as the first preset distance range; the preset distance range of the target obstacle with a speed greater than or equal to the speed of the vehicle is set as the second preset distance range. The lower limits of the first preset distance range and the second preset distance range are the same, and the upper limit of the first preset distance range is greater than the upper limit of the second preset distance, such as the first preset distance range is [0, 2], and the second preset distance range is [0, 1].

[0073] In the present invention, when there are multiple target key points whose lateral distances from the vehicle are within their corresponding preset distance ranges, the method further includes: determining the target obstacles corresponding to each of the multiple target key points as candidate main targets; determining the longitudinal distances between each candidate main target and the vehicle; and determining the candidate main target corresponding to the smallest longitudinal distance as the main target.

[0074] In an embodiment of the present invention, when there are multiple target obstacles whose respective lateral distances between their target key points and the vehicle are within their respective corresponding preset distance ranges, the multiple target obstacles are all determined as candidate main targets. Then, the longitudinal distances between each candidate main target and the vehicle are calculated. The candidate main target corresponding to the smallest longitudinal distance is determined as the main target.

[0075] For example, the target obstacles include target obstacle a, and the corresponding preset distance range is set to [0, 2], target obstacle b, and the corresponding preset distance range is set to [0, 1], target obstacle c and the corresponding preset distance range is set to [0, 1], target obstacle d and the corresponding preset distance range is set to [0, 1], and target obstacle e and the corresponding preset distance range is set to [0, 2]; then, the distance between the target key point of target obstacle a and the origin of the vehicle is determined to be 1.2, the distance between the target key point of target obstacle b and the origin of the vehicle is determined to be 0.8, the distance between the target key point of target obstacle c and the origin of the vehicle is determined to be 0.5, the distance between the target key point of target obstacle d and the origin of the vehicle is determined to be 1.5, and the distance between the target key point of target obstacle e and the origin of the vehicle is determined to be 2.5; thus, it is determined that the distances between the target obstacles a, b, c and the origin of the vehicle are within the corresponding preset distance ranges, and therefore the target obstacles a, b, c are all determined as candidate main targets. Determine the longitudinal distance between each candidate main target and the vehicle. Determine the longitudinal distance between candidate main target a and the vehicle to be 8m in front of the vehicle, determine the longitudinal distance between candidate main target b and the vehicle to be 10m in front of the vehicle, and determine the longitudinal distance between candidate main target c and the vehicle to be 20m in front of the vehicle. Therefore, candidate main target a, which has the smallest longitudinal distance value and is located 8m in front of the vehicle, is determined as the main target.

[0076] In the present invention, determining the longitudinal distance between each candidate main target and the vehicle includes: determining the first target key point that is closest to the vehicle in the longitudinal direction from the multiple key point coordinates of each candidate main target; and determining the longitudinal distance between the first target key point and the vehicle as the longitudinal distance between the candidate main target corresponding to the first target key point and the vehicle.

[0077] In an embodiment of the present invention, one implementation method for determining the longitudinal distance between each candidate main target and the vehicle is as follows: each candidate main target has four key point coordinates, and for any one of all the candidate main targets, a key point that is closest to the origin of the vehicle in the longitudinal direction is determined from the four key point coordinates of the any one of the main targets, and the key point is determined as the first target key point. Then, the longitudinal distance between the first target key point of each candidate main target and the origin of the vehicle is determined as the longitudinal distance between each candidate main target and the vehicle.

[0078] For example, the candidate main targets include candidate main targets a and b, and the first target key point of the candidate main target a is determined to be the key point coordinate a1 among the four key point coordinates, and the first target key point of the candidate main target b is determined to be the key point coordinate b1 among the four key point coordinates. Therefore, the longitudinal distance between the key point coordinate a1 and the origin of the vehicle is determined as the longitudinal distance between the candidate main target a and the vehicle, and the longitudinal distance between the key point coordinate b1 and the origin of the vehicle is determined as the longitudinal distance between the candidate main target b and the vehicle.

[0079] In the present invention, before determining the coordinates of multiple key points of the target obstacle based on the relative position coordinates, motion information and shape information of the target obstacle, the method also includes: determining whether the target obstacle has the intention to cut into the lane where the vehicle is located based on the motion information of the target obstacle; and deleting the target obstacle that does not have the intention to cut into the lane where the vehicle is located from the target obstacle queue.

[0080] In the embodiment of the present invention, the present invention finds that in the actual application scenario, although the distance between the obstacle and the self-vehicle is relatively close (the distance can be the lateral distance, the longitudinal distance, or a straight-line distance that considers both the lateral distance and the longitudinal distance), there is no intention to cut into the lane where the self-vehicle is located during the driving process. At this time, the self-vehicle can still drive forward safely without being affected by such obstacles. Therefore, in order to further reduce the consumption of computing resources, the present invention further determines whether each target obstacle has the intention to cut in after determining the target obstacle from the obstacles, and directly deletes the target obstacle without the intention to cut in from the target obstacle queue, and no longer participates in the execution of subsequent steps. Among them, one implementation method of determining whether the target obstacle has the intention to cut into the lane where the self-vehicle is located is: according to the heading angle of the target obstacle and the relative driving speed between the target obstacle and the self-vehicle, determine the time required for the body of the target obstacle to touch the x-axis of the self-vehicle coordinate system; when the time is lower than the set value, it is determined that the target obstacle has the intention to cut in. Among them, the set value can be set according to the actual application scenario, and is not specifically limited here. The x-axis direction of the ego vehicle coordinate system is the same as the driving direction of the ego vehicle. It should be understood that the implementation method of determining whether the target obstacle has the intention to cut into the lane where the ego vehicle is located can also be some other implementation methods, which are not specifically limited here.

[0081] In an embodiment of the present invention, Figure 4As shown in the figure, firstly, obstacle information is collected by the millimeter-wave radar and front camera of the ego vehicle, and the collected obstacle information is fused to obtain the target reference point of the obstacle. Then, based on the ego vehicle coordinate system constructed with the rear axle center of the ego vehicle as the origin, the relative position coordinates of the target reference point of each obstacle to the origin are calculated. Then, combined with the lane line information, the nearest obstacle in front of the lane where the ego vehicle is located is selected as the target obstacle, and the two obstacles in front of the ego vehicle in the left lane adjacent to the lane where the ego vehicle is located are selected as target obstacles, and the two obstacles in front of the ego vehicle in the right lane adjacent to the lane where the ego vehicle is located are selected as target obstacles, thereby obtaining five target obstacles. For the target obstacle in front of the lane where the ego vehicle is located, it is not necessary to participate in the judgment of whether there is an intention to cut in, but directly participate in the subsequent determination of the main target. For the target obstacle with the intention to cut in and the lane where the ego vehicle is located, the key point coordinates of each target obstacle are calculated based on the Bounding Box, and the target obstacle is determined as a frame target by the multiple key point coordinates of the target obstacle itself. Then, a key point closest to the vehicle is determined from the multiple key points of the target obstacle as the target key point, and a corresponding target key point can be determined for each target obstacle. After obtaining the target key points of each target obstacle, the lateral distances between the target key points of each target obstacle and the origin of the vehicle are calculated, and the calculated lateral distances are compared with the corresponding preset distance ranges. If there is a target key point whose lateral distance with the vehicle is within the corresponding preset distance range, the target obstacle corresponding to the target key point is determined as the main target; if there are multiple target key points whose lateral distances with the vehicle are within the corresponding preset distance ranges, the target obstacles corresponding to the multiple target key points are determined as candidate main targets, and then the longitudinal distances between each candidate main target and the vehicle are determined, and the candidate main target corresponding to the longitudinal distance with the smallest value is determined as the main target.

[0082] The present application also provides a straight cut target recognition device. Figure 5 As shown, a straight-line cut-in target recognition device 500 provided by the present application includes:

[0083] A target obstacle determination unit 501 is used to determine a target obstacle from the obstacles according to the relative position relationship between the ego vehicle and each obstacle;

[0084] A key point coordinate determining unit 502, configured to determine a plurality of key point coordinates of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle;

[0085] A target key point determination unit 503, configured to determine the target key point that is closest to the vehicle in the lateral direction from the coordinates of the multiple key points of the target obstacle;

[0086] A main target determination unit 504, configured to determine a main target from each target obstacle according to the determined target key points of each target obstacle;

[0087] The control unit 505 is used to control the vehicle to cruise based on the main target.

[0088] Optionally, the primary target determination unit 504 includes:

[0089] A lateral distance determination unit, configured to determine, based on the determined target key points of the respective target obstacles, the lateral distances between the respective target key points of the respective target obstacles and the ego vehicle;

[0090] A distance relationship determination unit, used to determine the relationship between the lateral distance between each target key point of each target obstacle and the vehicle and the corresponding preset distance range;

[0091] The first main target determination unit is used to determine a target obstacle corresponding to a target key point as a main target when the lateral distance between the target key point and the vehicle is within a preset distance range corresponding to the target key point.

[0092] Optionally, the device further comprises:

[0093] The preset distance range determining unit is used to determine the preset distance range corresponding to the target obstacle according to the relative speed between the target obstacle and the vehicle.

[0094] Optionally, the device further comprises:

[0095] A candidate main target determination unit is used to determine the target obstacles corresponding to each of the multiple target key points as candidate main targets when the lateral distances between the multiple target key points and the vehicle are within the corresponding preset distance ranges;

[0096] A longitudinal distance determination unit, used to determine the longitudinal distance between each candidate main target and the ego vehicle;

[0097] The second main target determining unit is used to determine the candidate main target corresponding to the smallest longitudinal distance as the main target.

[0098] Optionally, the longitudinal distance determination unit comprises:

[0099] A first target key point determination unit, configured to determine a first target key point that is closest to the vehicle in the longitudinal direction from the coordinates of the multiple key points of each candidate main target;

[0100] The longitudinal distance determination subunit is used to determine the longitudinal distance between the first target key point and the vehicle as the longitudinal distance between the candidate main target corresponding to the first target key point and the vehicle.

[0101] Optionally, the device further comprises:

[0102] A preset length determining unit, configured to set the length in the shape information of the target obstacle to a preset length when the length in the shape information of the target obstacle exceeds a first threshold;

[0103] The preset width determining unit is used to set the width in the shape information of the target obstacle to the preset width when the width in the shape information of the target obstacle exceeds a second threshold.

[0104] Optionally, the device further comprises:

[0105] A cutting-in intention determination unit, used to determine whether the target obstacle has the intention to cut into the lane where the vehicle is located according to the motion information of the target obstacle;

[0106] The filtering unit is used to delete the target obstacles that have no intention of cutting into the lane where the vehicle is located from the target obstacle queue.

[0107] The embodiment of the present application also provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus;

[0108] Memory, used to store computer programs;

[0109] The processor is used to implement the steps of the straight-line target identification method described in the first aspect of the present application when executing the program stored in the memory.

[0110] An embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of a straight-line cut-in target recognition method described in the first aspect of the present application are implemented.

[0111] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0112] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0113] The above description is only a preferred embodiment of the present application and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application are included in the protection scope of the present application.

Claims

1. A straight-line target recognition method, characterized in that: The method comprises: Determining a target obstacle from the obstacles according to the relative positional relationship between the vehicle and the obstacles; Determine the coordinates of multiple key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle; Determine the target key point that is closest to the vehicle in the lateral direction from the coordinates of the multiple key points of the target obstacle; Determining a main target from among the target obstacles according to the determined target key points of the target obstacles; The vehicle is controlled to cruise based on the main target.

2. The straight-line cut-in target recognition method according to claim 1, characterized in that: The determining of the main target from the target obstacles according to the determined target key points of the target obstacles includes: Determine the lateral distance between the target key points of each target obstacle and the ego vehicle according to the determined target key points of each target obstacle; Determine the relationship between the lateral distance between each target key point of each target obstacle and the ego vehicle and the corresponding preset distance range; When there is a target key point whose lateral distance to the vehicle is within the preset distance range corresponding to the vehicle, the target obstacle corresponding to the target key point is determined as the main target.

3. The straight-line cut-in target recognition method according to claim 2 is characterized in that: The method further comprises: According to the relative speed between the target obstacle and the vehicle, a preset distance range corresponding to the target obstacle is determined.

4. The straight-line cut-in target recognition method according to any one of claims 2 to 3, characterized in that: In the case where the lateral distances between a plurality of target key points and the vehicle are within the respective corresponding preset distance ranges, the method further includes: Determine the target obstacles corresponding to each of the plurality of target key points as candidate main targets; Determine the longitudinal distance between each candidate main target and the ego vehicle; The candidate main target corresponding to the smallest longitudinal distance is determined as the main target.

5. The straight-line cut-in target recognition method according to claim 4 is characterized in that: Determining the longitudinal distance between each candidate main target and the vehicle includes: Determine the first target key point that is closest to the vehicle in the longitudinal direction from the coordinates of the multiple key points of each candidate main target; The longitudinal distance between the first target key point and the vehicle is determined as the longitudinal distance between the candidate main target corresponding to the first target key point and the vehicle.

6. The straight-line cut-in target recognition method according to claim 1, characterized in that: The method further comprises: When the length in the shape information of the target obstacle exceeds a first threshold, setting the length in the shape information of the target obstacle to a preset length; When the width in the shape information of the target obstacle exceeds a second threshold, the width in the shape information of the target obstacle is set to a preset width.

7. The straight-line cut-in target recognition method according to claim 3 is characterized in that: Before determining the coordinates of a plurality of key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle, the method further includes: According to the motion information of the target obstacle, determine whether the target obstacle has the intention to cut into the lane where the ego vehicle is located; Target obstacles that do not intend to cut into the lane where the vehicle is located are deleted from the target obstacle queue.

8. A straight-line cut-in target recognition device, characterized in that: The device comprises: A target obstacle determination unit, configured to determine a target obstacle from among the obstacles according to a relative position relationship between the vehicle and each obstacle; A key point coordinate determination unit, used to determine the coordinates of a plurality of key points of the target obstacle according to the relative position coordinates, motion information and shape information of the target obstacle; A target key point determination unit, used to determine the target key point that is closest to the vehicle in the lateral direction from the coordinates of the multiple key points of the target obstacle; A main target determination unit, configured to determine a main target from each target obstacle according to the determined target key points of each target obstacle; A control unit is used to control the vehicle to cruise based on the main target.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor is used to implement the steps of the straight-line target identification method described in any one of claims 1-7 when executing the program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a straight-cut target recognition method as described in any one of claims 1 to 7 are implemented.

Images