Close-range obstacle avoidance method, device, vehicle, and storage medium based on laser radar

Abstract

The present invention discloses a close-range obstacle avoidance method, device, vehicle, and storage medium based on laser radar. The method comprises: obtaining image frames generated by a laser radar scanning the vehicle's surroundings in real time; upon detecting the presence of an obstacle within the vehicle's first obstacle avoidance zone in the current frame, sending a stop command to the vehicle, and recording the number of obstacles within the vehicle's second obstacle avoidance zone in the current frame, the number of which represents the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone; and upon detecting the absence of an obstacle within the first obstacle avoidance zone for N consecutive frames and the number of obstacles within the second obstacle avoidance zone being zero, sending a start command to the vehicle. The present invention relies solely on laser radar perception and observation information. If no obstacle is detected, the method achieves close-range obstacle avoidance by determining the number of obstacles within the vehicle's close-range zone.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a laser radar-based close-range obstacle avoidance method, device, vehicle, and storage medium. Background Art

[0002] With the development of AI technology, active obstacle avoidance technology is gaining increasing attention from engineers. It's a core component of autonomous vehicles, and its ability to accurately identify and detect obstacles and safely and accurately avoid them is a crucial skill.

[0003] Currently, in the field of autonomous driving, vehicle obstacle avoidance involves the safety, stability, and real-time performance of autonomous driving. If the multi-sensor fusion method cannot be used due to the on-site environment or other objective factors, it is relatively difficult to rely solely on the field-side lidar perception information for obstacle avoidance. Moreover, in close-range obstacle avoidance scenarios, due to the perception principle of lidar, the perception of people or bicycles will be integrated with the vehicle, resulting in the inability to perceive people through lidar. In addition, in the absence of vehicle-side ultrasonic assistance or auxiliary detection blind spots, the presence of people cannot be detected. Based on the above two points, the existing obstacle avoidance methods cannot guarantee the integrity and real-time performance of observation information due to the lack of radar perception information or the limitations of other sensors, resulting in the inability to identify people at close range, which in turn leads to the failure of obstacle avoidance and even safety accidents. Summary of the Invention

[0004] The technical problem to be solved by the embodiments of the present invention is to provide a close-range obstacle avoidance method, device, vehicle and storage medium based on laser radar, which only relies on laser radar perception and observation information. When no obstacles are perceived, close-range obstacle avoidance is achieved by judging the number of obstacles in the close range of the vehicle.

[0005] To achieve the above objectives, an embodiment of the present invention provides a close-range obstacle avoidance method based on laser radar, comprising:

[0006] Obtain image frames generated by the laser radar scanning the vehicle's surrounding information in real time;

[0007] When an obstacle is detected in a first obstacle avoidance zone of the vehicle in a current frame, a stop command is sent to the vehicle, and the number of obstacles in a second obstacle avoidance zone of the vehicle in the current frame is recorded, where the number represents the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone;

[0008] When it is detected for N consecutive frames that no obstacles exist in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is zero, a start instruction is sent to the vehicle.

[0009] As an improvement to the above solution, the first obstacle avoidance area is obtained by the following steps:

[0010] receiving in real time the location information sent by the vehicle; wherein the location information includes a horizontal coordinate, a vertical coordinate, and a direction;

[0011] Based on the position information, an area within a preset range of the vehicle is calculated in real time with the center of the vehicle as the geometric center to obtain a first obstacle avoidance area.

[0012] As an improvement to the above solution, after detecting that no obstacle exists in the first obstacle avoidance area for N consecutive frames, the method further includes:

[0013] The number of obstacles in the second obstacle avoidance area is updated according to the observation information in the image frame.

[0014] As an improvement to the above solution, updating the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame specifically includes:

[0015] Tracking observation information of obstacles within a preset range of the vehicle in M consecutive frames according to the principle of closest distance; wherein the observation information includes at least coordinates, speed, and direction;

[0016] Calculate whether the obstacle in each frame is within the second obstacle avoidance area according to the observation information;

[0017] The number of obstacles in the second obstacle avoidance area is increased or decreased according to the calculation result.

[0018] As an improvement to the above solution, increasing or decreasing the number of obstacles in the second obstacle avoidance area according to the calculation result specifically includes:

[0019] Based on the calculation results, a queue is used to record the status of each obstacle in M consecutive frames; wherein 1 indicates that the obstacle is in the second obstacle avoidance area in the current frame, and 0 indicates that the obstacle is not in the second obstacle avoidance area in the current frame;

[0020] Traversing the queue to determine whether the obstacle in the Mth frame is within the second obstacle avoidance area;

[0021] If the obstacle is within the second obstacle avoidance area, the number of obstacles within the second obstacle avoidance area increases accordingly;

[0022] If the obstacle is not in the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is reduced accordingly.

[0023] As an improvement to the above solution, the updating of the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame further includes:

[0024] When it is detected that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is not zero, the vehicle remains in a stopped state.

[0025] As an improvement to the above solution, a Hungarian tracking algorithm based on Mahalanobis distance is used when detecting obstacles in the image frame.

[0026] An embodiment of the present invention further provides a close-range obstacle avoidance device based on a laser radar, comprising:

[0027] An acquisition module is used to obtain image frames generated by the laser radar scanning the surrounding information of the vehicle in real time;

[0028] a parking module configured to, upon detecting the presence of an obstacle within a first obstacle avoidance zone of the vehicle in a current frame, send a parking instruction to the vehicle and record the number of obstacles within a second obstacle avoidance zone of the vehicle in the current frame, the number of obstacles representing the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone;

[0029] The starting module is configured to send a starting instruction to the vehicle when detecting for N consecutive frames that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is zero.

[0030] An embodiment of the present invention also provides a vehicle, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the laser radar-based close-range obstacle avoidance method described above is implemented.

[0031] An embodiment of the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any of the above-mentioned laser radar-based close-range obstacle avoidance methods.

[0032] Compared to the prior art, the embodiments of the present invention provide a lidar-based close-range obstacle avoidance method, device, vehicle, and storage medium with the following advantages: The method generates image frames by scanning the vehicle's surroundings in real time using a lidar; when an obstacle is detected within the vehicle's first obstacle avoidance zone in the current frame, a stop command is issued to the vehicle, and the number of obstacles within the vehicle's second obstacle avoidance zone in the current frame is recorded, indicating the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone; and when the absence of an obstacle in the first obstacle avoidance zone and the number of obstacles within the second obstacle avoidance zone are detected for N consecutive frames, a start command is issued to the vehicle. The embodiments of the present invention rely solely on lidar sensor observation information, determining the location of obstacles through frame-to-frame matching, and achieving close-range obstacle avoidance by determining the number of obstacles within the close-range zone. This ensures that obstacles are within a certain distance of the vehicle, ensuring that obstacle avoidance and parking are still effective despite the inability to sense them, thus ensuring the reliability and stability of obstacle avoidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a flow chart of a preferred embodiment of a close-range obstacle avoidance method based on laser radar provided by the present invention;

[0034] Figure 2 1 is a schematic diagram of detecting the position of an obstacle in a preferred embodiment of a close-range obstacle avoidance method based on a laser radar provided by the present invention;

[0035] Figure 3 1 is a schematic diagram of detecting the position of an obstacle in a preferred embodiment of a close-range obstacle avoidance method based on a laser radar provided by the present invention;

[0036] Figure 4 1 is a schematic diagram of detecting the position of an obstacle in a preferred embodiment of a close-range obstacle avoidance method based on a laser radar provided by the present invention;

[0037] Figure 5 1 is a schematic diagram of detecting the position of an obstacle in a preferred embodiment of a close-range obstacle avoidance method based on a laser radar provided by the present invention;

[0038] Figure 6 1 is a schematic diagram of detecting the position of an obstacle in a preferred embodiment of a close-range obstacle avoidance method based on a laser radar provided by the present invention;

[0039] Figure 7 1 is a schematic diagram of obstacle detection in a preferred embodiment of a close-range obstacle avoidance method based on laser radar provided by the present invention;

[0040] Figure 81 is a schematic structural diagram of a preferred embodiment of a close-range obstacle avoidance device based on laser radar provided by the present invention;

[0041] Figure 9 It is a structural schematic diagram of a preferred embodiment of a vehicle provided by the present invention. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0043] See also Figure 1 , Figure 1 The figure is a flow chart of a preferred embodiment of a laser radar-based close-range obstacle avoidance method provided by the present invention. The laser radar-based close-range obstacle avoidance method includes:

[0044] S1, obtains the image frame generated by the laser radar scanning the vehicle's surrounding information in real time;

[0045] S2, upon detecting the presence of an obstacle in a first obstacle avoidance zone of the vehicle in the current frame, sending a stop command to the vehicle, and recording the number of obstacles in a second obstacle avoidance zone of the vehicle in the current frame, the number of obstacles representing the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone;

[0046] S3: When it is detected for N consecutive frames that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is zero, a start instruction is sent to the vehicle.

[0047] Specifically, a field-side LiDAR (lidar) is used to scan the vehicle's perimeter, acquiring image frames generated by the real-time LiDAR scanning of the vehicle's perimeter. If an obstacle is detected within the vehicle's first obstacle avoidance zone in the current frame, and if the obstacle remains within the first obstacle avoidance zone for two consecutive frames, a stop command is issued to the vehicle, and the number of obstacles within the vehicle's second obstacle avoidance zone in the current frame is recorded. This number represents the dynamics of obstacles entering or leaving the second obstacle avoidance zone; the second obstacle avoidance zone is larger than the first. If the absence of an obstacle in the first obstacle avoidance zone is detected for N consecutive frames, this could be due to the absence of the obstacle in the first obstacle avoidance zone or the proximity of the obstacle to the LiDAR, preventing it from being detected. In this case, the number of obstacles in the second obstacle avoidance zone must be determined. If the absence of an obstacle in the first obstacle avoidance zone is detected, the number of obstacles in the second obstacle avoidance zone is zero, and if there are no obstacles in the second obstacle avoidance zone for five consecutive frames, this indicates that there are no obstacles around the vehicle, and a start command is issued to the vehicle.

[0048] It should be noted that N is a positive integer greater than 1. In this embodiment, N is preferably 5. That is, when no obstacles are detected in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is zero for five consecutive frames, a start command is sent to the vehicle. In actual application, N can be selected according to actual conditions.

[0049] This embodiment relies solely on LiDAR perception and observation information, and determines the position of obstacles through frame-to-frame matching. It also achieves close-range obstacle avoidance by determining the number of obstacles in the close-range area, thereby ensuring that obstacles are within a certain distance of the vehicle. Although perception cannot be obtained, the vehicle can still achieve the effect of obstacle avoidance and parking, thus ensuring the reliability and stability of obstacle avoidance.

[0050] In another preferred embodiment, the first obstacle avoidance area is obtained by the following steps:

[0051] receiving in real time the location information sent by the vehicle; wherein the location information includes a horizontal coordinate, a vertical coordinate, and a direction;

[0052] Based on the position information, an area within a preset range of the vehicle is calculated in real time with the center of the vehicle as the geometric center to obtain a first obstacle avoidance area.

[0053] Specifically, the system receives position information (pose(x, y, heading)) from the vehicle in real time, where x is the vehicle's horizontal coordinate, y is the vehicle's vertical coordinate, and heading is the vehicle's orientation. Based on this position information, the system calculates an area within a preset range of the vehicle in real time, with the center of the vehicle as the geometric center, to obtain a first obstacle avoidance zone. This calculation is performed at the frequency at which the vehicle sends position information.

[0054] It should be noted that the first obstacle avoidance area and the second obstacle avoidance area are not limited to rectangles and can be selected according to actual conditions.

[0055] In another preferred embodiment, after detecting that no obstacle exists in the first obstacle avoidance area for N consecutive frames, the method further includes:

[0056] The number of obstacles in the second obstacle avoidance area is updated according to the observation information in the image frame.

[0057] Specifically, if no obstacles are detected in the first obstacle avoidance zone for N consecutive frames, this could mean there are no obstacles in the first obstacle avoidance zone, or the obstacles may be too close to the vehicle to be detected by the LiDAR. In this case, the number of obstacles in the second obstacle avoidance zone needs to be updated based on the observation information in the image frames. By determining the number of obstacles in the second obstacle avoidance zone, close-range obstacle avoidance is achieved.

[0058] For example, see Figure 2-Figure 6 , Figure 2 、 Figure 3 、 Figure 4 、 Figure 5 、 Figure 6 This is a schematic diagram of obstacle detection in a preferred embodiment of a close-range obstacle avoidance method based on lidar provided by the present invention. The stop box represents the first obstacle avoidance area, the diff box represents the second obstacle avoidance area, and the diff box is one meter larger than the stop box in length and width. Pedestrians are obstacles. When an obstacle is detected in the first obstacle avoidance area in the current frame, the number of obstacles in the second obstacle avoidance area is recorded as the initial value of the difference between people entering and exiting the area. This is referred to as the initial value of in_out_dif. Figure 2 When pedestrian number one enters the stop frame, the vehicle stops and enters stop frame monitoring mode, detecting whether there are people in the stop frame in real time. If there are people, the vehicle stops. The number of pedestrians in the diff frame at this time is recorded as the initial value of in_out_dif (currently 4). Pedestrians number one and two are in the start_stop area. Figure 3 Pedestrian No. 1 walks out of the stop frame. At this time, there is only pedestrian No. 2 in the stop frame and it is still in the stop frame monitoring mode. At this time, the in_out_dif value is still 4. Figure 4 Pedestrians No. 1 and No. 4 walk out of the diff box, in_out_diff–2=2. At this time, there is only pedestrian No. 2 in the stop box, and the car is still in the stop box monitoring mode, so the car continues to stop. Figure 5 Pedestrian No. 3 walks out of the diff box, and in_out_diff – 1 = 1. Pedestrian No. 2 approaches the vehicle, and the radar perceives the pedestrian and vehicle as one, making the pedestrian invisible. At this point, there is no one in the stop box, and the system enters in_out_dif detection mode, which only detects the in_out_dif value. If it is not 0, the vehicle continues to stop. Figure 6 Pedestrian No. 2 walks out of the dif box, the in_out_diff value is 0, there is no one in the stop box, and the car leaves.

[0059] It should be noted that when the laser radar senses two close objects, it cannot distinguish the positions of the two objects and thinks they are one object. In view of this defect, when a person or a bicycle enters the area within about 30 cm of the vehicle, the radar cannot successfully sense the person and thinks that the person and the vehicle are one, resulting in the inability to detect people in the normal danger zone. The vehicle will continue to drive because there is no one around, but the person is still in the danger zone at this time, causing a safety accident. Based on the above, the in_out_dif monitoring mode is introduced to update the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame, and achieve close-range obstacle avoidance by judging the number of obstacles in the close-range area of the vehicle. The obstacles in this embodiment include but are not limited to pedestrians and bicycles, and the value of N can be selected according to actual needs.

[0060] In another preferred embodiment, updating the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame specifically includes:

[0061] Tracking observation information of obstacles within a preset range of the vehicle in M consecutive frames according to the principle of closest distance; wherein the observation information includes at least coordinates, speed, and direction;

[0062] Calculate whether the obstacle in each frame is within the second obstacle avoidance area according to the observation information;

[0063] The number of obstacles in the second obstacle avoidance area is increased or decreased according to the calculation result.

[0064] For example, when updating the number of obstacles within the second obstacle avoidance zone based on observation information in image frames, the vehicle tracks observation information for obstacles within 12 meters of the vehicle for four consecutive frames, using the closest distance principle. This observation information includes at least coordinates, speed, and direction. Based on this observation information, the vehicle calculates whether the obstacle is within the second obstacle avoidance zone in each frame. Based on the calculation result, the number of obstacles in the second obstacle avoidance zone is increased or decreased.

[0065] It should be noted that M is a positive integer greater than 1. In this embodiment, M is preferably 4, and the vehicle preset range is preferably within 12 meters of the vehicle. In actual application, M and the preset range can be selected according to actual conditions.

[0066] In another preferred embodiment, increasing or decreasing the number of obstacles in the second obstacle avoidance area according to the calculation result specifically includes:

[0067] Based on the calculation results, a queue is used to record the status of each obstacle in M consecutive frames; wherein 1 indicates that the obstacle is in the second obstacle avoidance area in the current frame, and 0 indicates that the obstacle is not in the second obstacle avoidance area in the current frame;

[0068] Traversing the queue to determine whether the obstacle in the Mth frame is within the second obstacle avoidance area;

[0069] If the obstacle is within the second obstacle avoidance area, the number of obstacles within the second obstacle avoidance area increases accordingly;

[0070] If the obstacle is not in the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is reduced accordingly.

[0071] Specifically, based on the calculation results, a queue is used to record the status of each obstacle in four consecutive frames. A value of 1 indicates that the obstacle is within the second obstacle avoidance area in the current frame, and a value of 0 indicates that the obstacle is not within the second obstacle avoidance area in the current frame. This queue is then traversed to determine whether the obstacle is within the second obstacle avoidance area in the fourth frame. If the obstacle is within the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is increased accordingly. If the obstacle is not within the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is decreased accordingly. For example, if the queue is 0 0 1 1, then in_out_dif + 1; if the queue is 1 1 0 0, then in_out_dif – 1.

[0072] In another preferred embodiment, the updating of the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame further includes:

[0073] When it is detected that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is not zero, the vehicle remains in a stopped state.

[0074] Specifically, after updating the number of obstacles in the second obstacle avoidance zone based on observations in the image frames, if no obstacles are detected in the first obstacle avoidance zone and the number of obstacles in the second obstacle avoidance zone is non-zero, it indicates that an obstacle is within close range of the vehicle. Therefore, the vehicle remains stopped. A start command is sent to the vehicle until the number of obstacles in the second obstacle avoidance zone reaches zero and there are no obstacles in the second obstacle avoidance zone for five consecutive frames.

[0075] As a preferred solution, when detecting obstacles in the image frame, a Hungarian tracking algorithm based on Mahalanobis distance is adopted.

[0076] For details, please refer to Figure 7 , Figure 7This is a schematic diagram of obstacle detection in a preferred embodiment of a close-range obstacle avoidance method based on laser radar provided by the present invention. Due to the characteristics of laser radar perception, when two pedestrians are close enough, they will be perceived as a bicycle or a motorcycle. Then when performing in_out_dif detection, if the matching between frames is performed based on Euclidean distance, the following situation will occur: in the Nth frame, pedestrian No. 1 is in the diff box, and pedestrian No. 3 is not far outside the diff box. At this time, pedestrian No. 1 walks towards pedestrian No. 3. In the N+1th frame, only the information of a bicycle outside the diff box can be obtained, so pedestrian No. 1 will lose the position information forever, resulting in only 1 in the queue detection of pedestrian No. 1, no 0, and in_out_dif cannot be set to 0, and the vehicle will be locked. Therefore, in order to solve the above problems, the Mahalanobis distance is introduced, and the speed of the observed pedestrian is added to the matching amount. In frame N, pedestrian 1's velocity is non-zero, but pedestrian 3's velocity is zero. In frame N+1, the bicycle's velocity and direction match those of pedestrian 1 in frame N through Hungarian matching based on Mahalanobis distance. Based on the Mahalanobis distance definition, we need to determine the covariance of the position and velocity of the same pedestrian between the two frames. Since motion models in real environments are not completely uniform, we assume that process noise originates from acceleration and add the acceleration noise transfer matrix to the uniform velocity model:

[0077]

[0078] v1=v0+aΔt

[0079] Right now

[0080]

[0081]

[0082] Among them, v0, d0, Δt are the velocity, displacement and time interval of the previous moment respectively, and v1, d1, Δt are the velocity, displacement and time interval of this moment respectively; x1=[d 1x ,d 1y ,v 1x ,v 1y ] T is the state at this moment, d 1x ,d 1y ,v 1x ,v 1y are the displacement and velocity in the x and y directions respectively; x0=[d 0x ,d 0y ,v 0x ,v 0y ] T is the state at the previous moment, d 0x ,d 0y ,v 0x ,v 0yare the displacement and velocity in the x and y directions respectively; a, a x ,a y is the acceleration, the acceleration in the x direction and the acceleration in the y direction.

[0083] make Since the defense difference matrix Q v =E(qq T )=GQG T Among them, the acceleration in the x direction is independent of the acceleration in the y direction. but

[0084]

[0085]

[0086] Correspondingly, the present invention also provides a close-range obstacle avoidance device based on laser radar, which can implement all the processes of the close-range obstacle avoidance method based on laser radar in the above embodiment.

[0087] See also Figure 8 , Figure 8 : This is a schematic structural diagram of a preferred embodiment of a laser radar-based short-range obstacle avoidance device provided by the present invention. The laser radar-based short-range obstacle avoidance device comprises:

[0088] An acquisition module 801 is used to acquire image frames generated by a laser radar scanning the surrounding information of the vehicle in real time;

[0089] a parking module 802 configured to, upon detecting the presence of an obstacle within the first obstacle avoidance zone of the vehicle in the current frame, send a parking instruction to the vehicle and record the number of obstacles within the second obstacle avoidance zone of the vehicle in the current frame, the number of obstacles representing the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone;

[0090] The starting module 803 is configured to send a starting instruction to the vehicle when it is detected for N consecutive frames that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is zero.

[0091] Preferably, the first obstacle avoidance area is obtained by the following steps:

[0092] receiving in real time the location information sent by the vehicle; wherein the location information includes a horizontal coordinate, a vertical coordinate, and a direction;

[0093] Based on the position information, an area within a preset range of the vehicle is calculated in real time with the center of the vehicle as the geometric center to obtain a first obstacle avoidance area.

[0094] Preferably, after the starting module 803 detects that no obstacle exists in the first obstacle avoidance area for N consecutive frames, the process further includes:

[0095] The number of obstacles in the second obstacle avoidance area is updated according to the observation information in the image frame.

[0096] Preferably, updating the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame specifically includes:

[0097] Tracking observation information of obstacles within a preset range of the vehicle in M consecutive frames according to the principle of closest distance; wherein the observation information includes at least coordinates, speed, and direction;

[0098] Calculate whether the obstacle in each frame is within the second obstacle avoidance area according to the observation information;

[0099] The number of obstacles in the second obstacle avoidance area is increased or decreased according to the calculation result.

[0100] Preferably, increasing or decreasing the number of obstacles in the second obstacle avoidance area according to the calculation result specifically includes:

[0101] Based on the calculation results, a queue is used to record the status of each obstacle in M consecutive frames; wherein 1 indicates that the obstacle is in the second obstacle avoidance area in the current frame, and 0 indicates that the obstacle is not in the second obstacle avoidance area in the current frame;

[0102] Traversing the queue to determine whether the obstacle in the Mth frame is within the second obstacle avoidance area;

[0103] If the obstacle is within the second obstacle avoidance area, the number of obstacles within the second obstacle avoidance area increases accordingly;

[0104] If the obstacle is not in the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is reduced accordingly.

[0105] Preferably, the updating of the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame further includes:

[0106] When it is detected that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is not zero, the vehicle remains in a stopped state.

[0107] Preferably, when detecting obstacles in the image frame, a Hungarian tracking algorithm based on Mahalanobis distance is adopted.

[0108] In specific implementation, the working principle, control process and technical effect of the close-range obstacle avoidance device based on laser radar provided in the embodiment of the present invention are the same as those of the close-range obstacle avoidance method based on laser radar in the above embodiment, and will not be repeated here.

[0109] See also Figure 9 , Figure 9 is a schematic structural diagram of a preferred embodiment of a vehicle provided by the present invention. The vehicle includes a processor 901, a memory 902, and a computer program stored in the memory 902 and configured to be executed by the processor 901. When the processor 901 executes the computer program, it implements the laser radar-based close-range obstacle avoidance method described in any of the above embodiments.

[0110] Preferably, the computer program can be divided into one or more modules / units (e.g., computer program 1, computer program 2, ...). These one or more modules / units are stored in the memory 902 and executed by the processor 901 to implement the present invention. The one or more modules / units can be a series of computer program instruction segments capable of implementing specific functions, and these instruction segments are used to describe the execution process of the computer program in the vehicle.

[0111] The processor 901 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor, or the processor 901 can be any conventional processor. The processor 901 is the control center of the vehicle, and uses various interfaces and lines to connect various parts of the vehicle.

[0112] The memory 902 mainly includes a program storage area and a data storage area. The program storage area can store an operating system, at least one application required for a function, and the data storage area can store related data. In addition, the memory 902 can be a high-speed random access memory or a non-volatile memory, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, and a flash card. Alternatively, the memory 902 can be other volatile solid-state memory devices.

[0113] It should be noted that the above vehicle may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that Figure 9 The structural diagram is merely an example of the above-mentioned vehicle and does not constitute a limitation on the above-mentioned vehicle. It may include more or fewer components than shown in the figure, or combine certain components, or different components.

[0114] An embodiment of the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the laser radar-based close-range obstacle avoidance method described in any of the above embodiments.

[0115] Embodiments of the present invention provide a laser radar-based close-range obstacle avoidance method, device, vehicle, and storage medium. The method obtains image frames generated by real-time laser radar scanning of the vehicle's surroundings. Upon detecting the presence of an obstacle within the vehicle's first obstacle avoidance zone in the current frame, a stop command is issued to the vehicle, and the number of obstacles within the vehicle's second obstacle avoidance zone in the current frame is recorded, indicating the dynamics of obstacles entering or leaving the second obstacle avoidance zone. The second obstacle avoidance zone is larger than the first obstacle avoidance zone. Upon detecting the absence of an obstacle within the first obstacle avoidance zone for N consecutive frames, and the number of obstacles within the second obstacle avoidance zone being zero, a start command is issued to the vehicle. This embodiment of the present invention relies solely on laser radar sensor observation information, determining the location of obstacles through frame-to-frame matching. It also achieves close-range obstacle avoidance by determining the number of obstacles within the close-range zone. This ensures that obstacles within a certain distance of the vehicle can still achieve obstacle avoidance and parking, even without sensory acquisition, ensuring the reliability and stability of obstacle avoidance.

[0116] It should be noted that the system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the system embodiment provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive work.

[0117] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A close-range obstacle avoidance method based on laser radar, characterized in that: include: Obtain image frames generated by the laser radar scanning the vehicle's surrounding information in real time; When an obstacle is detected in a first obstacle avoidance zone of the vehicle in a current frame, a stop command is sent to the vehicle, and the number of obstacles in a second obstacle avoidance zone of the vehicle in the current frame is recorded, where the number represents the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone; After N consecutive frames have detected that no obstacles are present within the first obstacle avoidance area, the observation information of obstacles within the preset range of the vehicle is tracked for M consecutive frames based on the principle of closest distance, wherein the observation information includes at least coordinates, speed, and direction; and whether the obstacle is within the second obstacle avoidance area in each frame is calculated based on the observation information. The number of obstacles in the second obstacle avoidance area is increased or decreased according to the calculation result, and the state of each obstacle in M consecutive frames is recorded in a queue based on the calculation result; where 1 indicates that the obstacle is in the second obstacle avoidance area in the current frame, and 0 indicates that the obstacle is not in the second obstacle avoidance area in the current frame; When the number of obstacles in the second obstacle avoidance area is zero, sending a start command to the vehicle; When detecting obstacles in the image frame, a Hungarian tracking algorithm based on Mahalanobis distance is adopted.

2. The laser radar-based close-range obstacle avoidance method according to claim 1, characterized in that: The first obstacle avoidance area is obtained by the following steps: receiving position information sent by the vehicle in real time; wherein the position information includes a horizontal coordinate, a vertical coordinate and a direction; based on the position information, taking the center of the vehicle as the geometric center, calculating in real time the area within the preset range of the vehicle to obtain the first obstacle avoidance area.

3. The close-range obstacle avoidance method based on laser radar according to claim 1, characterized in that: Increasing or decreasing the number of obstacles in the second obstacle avoidance area based on the calculation result specifically includes: traversing the queue to determine whether the obstacle in the Mth frame is in the second obstacle avoidance area; if the obstacle is in the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is increased accordingly; if the obstacle is not in the second obstacle avoidance area, the number of obstacles in the second obstacle avoidance area is decreased accordingly.

4. The laser radar-based close-range obstacle avoidance method according to claim 1, wherein: The method further includes updating the number of obstacles in the second obstacle avoidance area according to the observation information in the image frame, and then further includes: when it is detected that there are no obstacles in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is not zero, the vehicle remains in a stopped state.

5. A close-range obstacle avoidance device based on laser radar, characterized in that: include: An acquisition module is used to obtain image frames generated by the laser radar scanning the surrounding information of the vehicle in real time; a parking module configured to, upon detecting the presence of an obstacle within a first obstacle avoidance zone of the vehicle in a current frame, send a parking command to the vehicle and record the number of obstacles within a second obstacle avoidance zone of the vehicle in the current frame, the number of obstacles representing the dynamics of obstacles entering or leaving the second obstacle avoidance zone; wherein the second obstacle avoidance zone is larger than the first obstacle avoidance zone; and wherein a Hungarian tracking algorithm based on Mahalanobis distance is used to detect obstacles in the image frame; A starting module is configured to send a starting instruction to the vehicle when it is detected in N consecutive frames that there is no obstacle in the first obstacle avoidance area and the number of obstacles in the second obstacle avoidance area is zero; specifically, when it is detected in N consecutive frames that there is no obstacle in the first obstacle avoidance area, the observation information of obstacles within a preset range of the vehicle is tracked in M consecutive frames according to the principle of closest distance; wherein the observation information includes at least coordinates, speed and direction; calculate whether the obstacle in each frame is within the second obstacle avoidance area according to the observation information; increase or decrease the number of obstacles in the second obstacle avoidance area according to the calculation result, and use a queue to record the status of each obstacle in M consecutive frames based on the calculation result; wherein 1 indicates that the obstacle in the current frame is within the second obstacle avoidance area, and 0 indicates that the obstacle in the current frame is not within the second obstacle avoidance area; when the number of obstacles in the second obstacle avoidance area is zero, a starting instruction is sent to the vehicle.

6. A vehicle, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the close-range obstacle avoidance method based on laser radar as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein, when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the laser radar-based close-range obstacle avoidance method according to any one of claims 1 to 4.

Images