Method, device, equipment and medium for detecting laser radar mirror surface reflection ghost

Abstract

The application discloses a kind of detection method, device, equipment and medium of laser radar mirror reflection ghosting.It is characterized by including: obtaining the obstacle information corresponding to multiple target obstacles collected by laser radar in target vehicle;For the target obstacle to be detected currently, the mirror reflection ghosting probability corresponding to the target obstacle is determined according to the obstacle information associated with the target obstacle, wherein the mirror reflection ghosting probability is used to represent the probability that the target obstacle is mirror reflection ghosting;If the mirror reflection ghosting probability of the target obstacle is greater than the preset ghosting probability threshold, the target obstacle is determined as mirror reflection ghosting.It realizes the accurate detection of mirror reflection ghosting collected by vehicle laser radar in driving state, improves the identification accuracy of vehicle laser radar to obstacle, improves the automatic driving performance of vehicle, and ensures the safety of vehicle automatic driving.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving, and in particular to a method, apparatus, device, and medium for detecting ghost images reflected from lidar mirrors. Background Technology

[0002] Autonomous driving is a crucial direction in current vehicle development. Through its autonomous driving functions, vehicles can automatically identify obstacle and road information, enabling safe autonomous driving. Obstacle identification primarily relies on various sensors within the vehicle, such as LiDAR. LiDAR actively emits infrared light to detect target distances, thereby reconstructing the point cloud environment of the road where the vehicle is located, ensuring the safe operation of autonomous driving functions. However, in the current LiDAR detection process, due to the reflective properties of light and mirror-like objects on the road, specular reflections can occur in the direction the LiDAR detects obstacles, creating redundant obstacle information and affecting the performance of autonomous driving. Existing technologies typically perform clustering processing on the point cloud data from LiDAR, often failing to identify these specular reflections within the point cloud environment. Summary of the Invention

[0003] This invention provides a method, apparatus, device, and medium for detecting ghost images reflected from lidar mirrors, so as to achieve accurate detection of ghost images reflected from lidar mirrors.

[0004] According to one aspect of the present invention, a method for detecting ghost images from mirror reflections in lidar is provided, comprising:

[0005] Obtain obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle;

[0006] For the target obstacle to be detected, the probability of a specular reflection ghost image corresponding to the target obstacle is determined based on the obstacle information associated with the target obstacle, wherein the specular reflection ghost image probability is used to represent the probability that the target obstacle is a specular reflection ghost image;

[0007] If the probability of a specular reflection ghost image of the target obstacle is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular reflection ghost image.

[0008] According to another aspect of the present invention, a device for detecting ghost images of laser radar mirror reflections is provided, comprising:

[0009] The lidar data transmission module is used to acquire obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle;

[0010] The specular reflection ghost detection module is used to determine the specular reflection ghost probability corresponding to the target obstacle based on the obstacle information associated with the target obstacle, wherein the specular reflection ghost probability is used to represent the probability that the target obstacle is a specular reflection ghost.

[0011] The specular reflection ghost detection module is used to determine the target obstacle as a specular reflection ghost if the probability of the specular reflection ghost of the target obstacle is greater than a preset ghost probability threshold.

[0012] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0013] At least one processor; and

[0014] A memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the laser radar specular reflection ghost detection method according to any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the laser radar specular reflection ghost detection method according to any embodiment of the present invention.

[0017] The technical solution of this invention improves the efficiency of obstacle information collection by acquiring obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle, thereby increasing the probability of identifying specular reflection ghost images. For the target obstacle to be detected, the probability of a specular reflection ghost image corresponding to the target obstacle is determined based on the obstacle information associated with the target obstacle. The specular reflection ghost image probability represents the probability that the target obstacle is a specular reflection ghost image. For the target obstacle, the obstacle information is used to identify the target obstacle, determine the characteristics of the target obstacle, and compare... This method analyzes the probability of a target obstacle being a specular ghost image, thereby increasing the probability of identifying specular ghost images. If the probability of a target obstacle being a specular ghost image is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular ghost image. When the probability of a target obstacle being a specular ghost image meets the threshold requirement, it is identified as a specular ghost image, reducing the error rate of obstacle identification and further improving the accuracy of specular ghost image identification. This achieves accurate identification of specular ghost images around the target vehicle, solving the technical problem of not being able to identify specular ghost images in existing LiDAR technology, improving the autonomous driving performance of the vehicle, and ensuring the safety of autonomous driving.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of a method for detecting ghost images from mirror reflections of a lidar device, provided in Embodiment 1 of the present invention.

[0021] Figure 2 A schematic diagram simulating the process of specular reflection ghosting provided in an embodiment of the present invention;

[0022] Figure 3 This is a flowchart of another method for detecting ghost images from laser radar mirror reflections provided in Embodiment 2 of the present invention;

[0023] Figure 4 A flowchart of another method for detecting ghost images from mirror reflections in lidar provided in an embodiment of the present invention;

[0024] Figure 5 This is a schematic diagram of the structure of a laser radar mirror reflection ghost detection device provided in Embodiment 3 of the present invention;

[0025] Figure 6 This is a schematic diagram of the structure of an electronic device that implements the laser radar specular reflection ghost detection method according to embodiments of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] Example 1

[0028] Figure 1 This is a flowchart of a method for detecting ghost images from a lidar mirror reflection, provided in Embodiment 1 of the present invention. This embodiment is applicable to detecting ghost images from a target vehicle during its movement. The method can be executed by a lidar mirror ghost image detection device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0029] S110. Obtain obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle.

[0030] The target vehicle can be a vehicle with autonomous driving capabilities. The target vehicle is equipped with at least one LiDAR. It should be noted that the LiDAR is typically installed on the front, rear, and multiple sides of the roof of the target vehicle. The LiDAR scans the three-dimensional information of the surrounding environment and obstacles, and calculates the distance, orientation, height, and speed of each obstacle around the vehicle.

[0031] The target obstacle can be an obstacle detected by the target vehicle using lidar. The target obstacle is located during the target vehicle's movement and has the potential to obstruct the vehicle's movement. It typically consists of buildings, vehicles, pedestrians, animals, and other road infrastructure surrounding the vehicle.

[0032] The obstacle information can include the coordinates, size, point cloud, velocity, and acceleration information of the obstacle detected by the lidar.

[0033] Specifically, after the target vehicle starts, the lidar installed in the vehicle collects data around the vehicle. Based on the rays reflected from surrounding obstacles, it determines the obstacle information corresponding to each obstacle and stores the obstacles and their corresponding information in appropriate storage devices. The target vehicle then retrieves the multiple target obstacles and their corresponding information collected by the lidar from the storage devices.

[0034] S120. For the target obstacle to be detected, determine the probability of the specular reflection ghost image corresponding to the target obstacle based on the obstacle information associated with the target obstacle.

[0035] The specular reflection ghost probability is used to represent the probability that the target obstacle is a specular reflection ghost.

[0036] Specifically, the process begins by identifying the target obstacle to be inspected. Based on the obstacle information corresponding to the target obstacle, the probability that the target obstacle is a specular reflection ghost is determined.

[0037] Optionally, after the lidar emits its detection light, due to the presence of smooth mirror surfaces around the target vehicle, the lidar's emitted light, reflected by these mirrors, strikes the highly reflective target and reflects back, forming a specular ghost image. Because the light's travel time is longer, the position of this specular ghost image is usually on the extension line of the light source and the smooth mirror surface. Therefore, the position of the specular ghost image has an occlusion relationship with the smooth mirror surface. When the target obstruction is blocked by other obstacles in the same direction, the target obstacle has a certain probability of forming a specular ghost image. The highly reflective target can be a high-reflection sign, which is usually made of highly transparent reflective material and can reflect various types of light, such as lidar emitted light and sunlight.

[0038] For example, Figure 2 This is a schematic diagram simulating the process of specular reflection ghosting provided in an embodiment of the present invention. Figure 2 As shown: The target vehicle is driving on the road. The smooth mirror surface is provided by other motor vehicles. The laser radar of the target vehicle emits detection light, which is reflected at the ideal point of other motor vehicles and hits the high-reflection sign. The ghost point of the mirror reflection ghost is formed on the extension line of the target vehicle and other motor vehicles.

[0039] Optionally, the formation of specular reflection ghost images is related to high-reflectivity targets around the target vehicle. Since specular reflection ghost images are formed when the emitted light from the lidar is reflected by a smooth mirror onto a high-reflectivity target, the smooth mirror and the high-reflectivity target do not experience significant energy loss during the formation process. Therefore, high-reflectivity points must exist within the specular reflection ghost image. When a large number of high-reflectivity points exist in the point cloud corresponding to the target obstruction, there is a certain probability that the target obstacle is a specular reflection ghost image.

[0040] Optionally, during the target vehicle's movement, since the specular reflection ghost image is formed by the reflection of a high-reflection target through a smooth mirror, the specular reflection ghost image and the high-reflection target have the same depth. The movement trajectory of the specular reflection ghost image differs from that of other obstacles. The specular reflection ghost image will exhibit lateral movement as the relative position of the smooth mirror and the lidar changes, while normal obstacles will move along the road. This means that the movement trajectory of the specular reflection ghost image intersects with the trajectories of other obstacles around the target vehicle. When an abnormal movement trajectory of a target obstacle is detected, there is a certain probability that the target obstacle is a specular reflection ghost image.

[0041] S130. If the probability of the specular reflection ghost image of the target obstacle is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular reflection ghost image.

[0042] The preset ghost probability threshold can be a pre-set probability threshold used to determine whether a target obstacle is a specular reflection ghost.

[0043] Specifically, after determining the specular ghosting probability of the target obstacle, a preset ghosting probability threshold is obtained. The relationship between the specular ghosting probability of the target obstacle and the preset ghosting probability threshold is determined. If the specular ghosting probability of the target obstacle is greater than the preset ghosting probability threshold, the target obstacle is identified as a specular ghosting obstacle; if the specular ghosting probability of the target obstacle is not greater than the preset ghosting probability threshold, the target obstacle is identified as a normal obstacle.

[0044] The technical solution of this invention improves the efficiency of obstacle information collection by acquiring obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle, thereby increasing the probability of identifying specular reflection ghost images. For the target obstacle to be detected, the probability of a specular reflection ghost image corresponding to the target obstacle is determined based on the obstacle information associated with the target obstacle. The specular reflection ghost image probability represents the probability that the target obstacle is a specular reflection ghost image. For the target obstacle, the obstacle information is used to identify the target obstacle, determine the characteristics of the target obstacle, and compare... This method analyzes the probability of a target obstacle being a specular ghost image, thereby increasing the probability of identifying specular ghost images. If the probability of a target obstacle being a specular ghost image is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular ghost image. When the probability of a target obstacle being a specular ghost image meets the threshold requirement, it is identified as a specular ghost image, reducing the error rate of obstacle identification and further improving the accuracy of specular ghost image identification. This achieves accurate identification of specular ghost images around the target vehicle, solving the technical problem of not being able to identify specular ghost images in existing LiDAR technology, improving the autonomous driving performance of the vehicle, and ensuring the safety of autonomous driving.

[0045] Example 2

[0046] Figure 3 This is a flowchart of another method for detecting specular reflection ghosting using lidar, provided in Embodiment 2 of the present invention. The relationship between this embodiment and the above embodiments is that this is a specific method for determining the probability of specular reflection ghosting corresponding to a target obstacle. For example... Figure 3 As shown, the method for detecting ghost images reflected from the lidar mirror includes:

[0047] S310: Obtain obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle.

[0048] Optionally, in another optional embodiment of the present invention, the step of obtaining obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle further includes: obtaining the specular reflection ghosting filtering range of the target vehicle; and determining the target obstacles of the target vehicle based on the obstacle position information and the specular reflection ghosting filtering range.

[0049] The specular reflection ghosting filtering range can be used to filter obstacles detected by the LiDAR on a target vehicle. This filtering range can be pre-calibrated and stored within the target vehicle by a calibration engineer. The filtering range can be defined with the target vehicle as the origin, with the lower and upper boundaries in the x-direction and the upper and lower boundaries in the y-direction. It should be noted that during vehicle operation, to ensure safe driving, the target vehicle's trajectory is typically automatically adjusted based on obstacles. Since specular reflection ghosting often appears within a certain range in front of the target vehicle, it affects driving safety. Therefore, to improve the filtering efficiency of specular reflection ghosting, the filtering range is calibrated to enhance detection efficiency.

[0050] The obstacle location information can be the center coordinates of the obstacle obtained by LiDAR. The obstacle location information includes the three-dimensional coordinates of the center point.

[0051] Specifically, the system obtains the target vehicle's calibrated specular reflection ghosting filtering range and obstacle location information, filters obstacles around the target vehicle based on the specular reflection ghosting filtering range and obstacle location information, and determines the target obstacles within the specular reflection ghosting filtering range.

[0052] For example, the obstacle location information detected by the target vehicle can be represented by "center(x,y,z)". The filtering range of the specular reflection ghost is based on the target vehicle as the origin of the coordinate system. The lower and upper boundaries of the filtering in the x-direction are represented by "select_x_min and select_x_max" respectively, and the upper and lower boundaries of the filtering in the y-direction are represented by "select_y_min and select_y_max" respectively. The filtering range of the specular reflection ghost can be expressed as:

[0053] select_x_min <center.x<select_x_max

[0054] select_y_min <center.y<select_y_max

[0055] Optionally, the obstacles around the target vehicle are filtered by the ghosting filtering range and obstacle position information through mirror reflection to obtain multiple filtered target obstacles. The multiple target obstacles are stored in the candidate ghosting list in sequence, and the obstacle information corresponding to each target obstacle is maintained.

[0056] S320. Calculate the occlusion ratio, mirror reflection ratio, and velocity feature probability of the target obstacle based on the obstacle information associated with the target obstacle.

[0057] Here, occlusion ratio can be the proportion of the target object that is occluded by other obstacles in the same direction. Specular high-reflection ratio can be the ratio between high-reflection points in the target obstacle and the point cloud data contained in the target obstacle. Velocity feature probability can be the probability that the target obstacle's trajectory is abnormal.

[0058] Specifically, for the obstacle information associated with the target obstacle, the occlusion ratio of the target obstacle being occluded by other obstacles in the same direction is calculated; the mirror reflection ratio of the target obstacle is calculated using the point cloud data corresponding to the target obstacle; the motion trajectory of the target obstacle relative to the target vehicle is determined by the speed and acceleration of the target obstacle, and then the speed feature probability of the target obstacle is determined.

[0059] Optionally, in another optional embodiment of the present invention, the obstacle information includes obstacle location information, obstacle point cloud index information, obstacle size information, and obstacle velocity information; the step of calculating the occlusion ratio, specular reflection ratio, and velocity feature probability of the target obstacle based on the obstacle information includes: determining the occlusion ratio of the target obstacle to other obstacles based on the obstacle location information and the obstacle size information; determining the specular reflection ratio of the target obstacle based on the obstacle point cloud index information; and determining the velocity feature probability of the target obstacle based on the obstacle velocity information and the obstacle distance information.

[0060] The obstacle point cloud index information can be the index information of the target vehicle's LiDAR-acquired point cloud data associated with the corresponding target obstacle. When the LiDAR scans the area around the target vehicle, corresponding index information is set for each obstacle to associate with the point cloud data of each obstacle. The obstacle point cloud index information can be stored in a point cloud data ID list corresponding to each obstacle, where the point cloud data ID is used to uniquely identify a point cloud data. For example, the point cloud data ID list can be represented by "instance_id", and the obstacle point cloud index information can be represented by "index".

[0061] The obstacle size information can be the span in the arc direction and the span in the opposite height direction in a polar coordinate system centered on the target vehicle. For example, in the obstacle size information, the span in the arc direction can be represented by "rad_min, rad_max", and the span in the opposite height direction can be represented by "height_min, height_max".

[0062] Among them, obstacle speed information can be the speed and acceleration of the target obstacle in the x and y directions respectively in a coordinate system centered on the target vehicle.

[0063] Specifically, the system acquires the obstacle location and size information of the target obstacle, calculates the occlusion ratio between the target obstacle and other obstacles in the same direction, acquires the obstacle point cloud index information of the target obstacle, calculates the mirror reflection ratio of the target obstacle, acquires the obstacle speed and distance information of the target obstacle, determines the motion trajectory of the target obstacle, and calculates the speed feature probability of the target obstacle.

[0064] Optionally, in another optional embodiment of the present invention, determining the occlusion ratio of the target obstacle to other obstacles based on the obstacle location information and the obstacle size information includes:

[0065] Calculate the distance relationship between the target obstacle and other obstacles based on the obstacle location information; determine the occlusion ratio between the target obstacle and other obstacles based on the distance relationship and the obstacle size information.

[0066] The distance relationship can refer to the distance between the target obstacle and other obstacles and the target vehicle. It should be noted that since specular reflection ghosts are usually located far from other obstacles in the same direction, the distance relationship between the target obstacle and other obstacles is calculated to determine the other obstacles closest to the target obstacle, at which point those other obstacles best obscure the target obstacle.

[0067] Optionally, the system retrieves the obstacle location information corresponding to each target obstacle in the ghost candidate list, determines the distance between each target obstacle and the target vehicle, and sets a distance sorting list, arranging the obstacles according to the magnitude of their distances to the target vehicle. Then, based on the distance sorting list, the target obstacle closest to the target obstacle is determined as the nearest other obstacle.

[0068] Optionally, after determining the nearest other obstacle to the target obstacle based on the distance relationship between the target obstacle and other obstacles, obtain the obstacle size information of the nearest other obstacle, and calculate the occlusion ratio between the obstacles in the arc direction and the height direction based on the span of the target obstacle in the arc direction and the span of the nearest other obstacle in the arc direction and the height direction in the polar coordinate system.

[0069] For example, the span of the target obstacle in the arc direction and the span in the height direction in the polar coordinate system are represented by "height_min1 and height_max1" respectively, and the nearest other obstacles are represented by "height_min2 and height_max2" respectively. The occlusion ratio between obstacles in the arc direction and the height direction is calculated as "covered_ratio_rad and covered_ratio_height".

[0070] Optionally, in another optional embodiment of the present invention, determining the specular reflection ratio of the target obstacle based on the obstacle point cloud index information and the obstacle size information includes:

[0071] Acquire the raw point cloud information detected by the lidar; determine the specular reflectance of the target obstacle and the number of obstacle point clouds based on the raw point cloud information and the obstacle point cloud index information; determine the specular reflectance ratio of the target obstacle based on the specular reflectance and the number of obstacle point clouds.

[0072] The raw point cloud information can be point cloud information collected by the target vehicle's LiDAR; the raw point cloud information records the energy intensity of each point cloud. It should be noted that the raw point cloud information has corresponding obstacle point cloud index information in the point cloud data ID list, and the obstacle to which each raw point cloud belongs can be identified through the index information.

[0073] The number of highly reflective points can be the number of highly reflective points in the original point cloud information corresponding to the target obstacle. Optionally, highly reflective points can be points whose energy intensity is greater than a preset point cloud energy intensity threshold.

[0074] The number of obstacle point clouds can be the number of original point clouds corresponding to the target obstacle.

[0075] Specifically, the process involves acquiring the raw point cloud information detected by the target vehicle's lidar, determining the number of obstacle point clouds corresponding to the target obstacle based on the obstacle point cloud index information, determining the energy intensity of the point cloud corresponding to the target obstacle based on the raw point cloud information, and sequentially determining whether the energy intensity of the point cloud is greater than a preset point cloud energy intensity threshold. If the energy intensity of the point cloud is greater than the preset point cloud energy intensity threshold, it is considered a high-reflection point, and the number of high-reflection points is determined. The ratio of the number of high-reflection points to the number of obstacle point clouds is calculated to obtain the high-reflection ratio of the target obstacle.

[0076] For example, the number of obstacle point clouds corresponding to the target obstacle is represented by "n", the number of specular reflections is represented by "high_i_num", and the specular reflection ratio is represented by "high_intensity_ratio". The specular reflection ratio is calculated as follows:

[0077] high_intensity_ratio=high_i_num / n

[0078] Optionally, in another optional embodiment of the present invention, determining the velocity feature probability of the target obstacle based on the obstacle velocity information and the obstacle distance information includes: determining candidate obstacles of the target obstacle based on the obstacle distance information; calculating the occlusion radian angle of the target obstacle based on the obstacle velocity information of the target obstacle and the obstacle velocity information of the candidate obstacles; and determining the velocity feature probability of the target obstacle based on the occlusion radian angle.

[0079] The occlusion arc angle can be the arc angle of occlusion between the trajectory of the target obstacle and the trajectory of the candidate obstacle. It should be noted that after calculating the trajectory of the target obstacle relative to the target vehicle and the trajectory of the candidate obstacle relative to the target vehicle, the occlusion arc angle between them is determined based on their trajectories. When the trajectory of the specular reflection ghost image relative to the target vehicle and the smooth mirror surface exhibits lateral movement, the relative angle between the trajectory of the specular reflection ghost image and the target vehicle and the smooth mirror surface is 90°. Therefore, the velocity characteristic probability of the target obstacle is determined based on the occlusion arc angle.

[0080] Candidate obstacles can be other obstacles that are closest to the target obstacle.

[0081] Optionally, obstacle velocity information includes the obstacle's velocity and acceleration in the x and y directions. Based on obstacle distance information, the candidate obstacle closest to the target obstacle is determined. Obstacle velocity information of the target obstacle and the candidate obstacle is obtained separately. When the velocity of the target obstacle and the candidate obstacle in the x and y directions is not zero, the occlusion radian angle of the target obstacle is determined based on the distance to the target obstacle and the velocity of the candidate obstacle in the x and y directions. The velocity feature probability of the target obstacle is determined based on the occlusion radian angle.

[0082] For example, the velocities of the candidate obstacle in the x and y directions are vx1 and vy1, respectively; the velocities of the target obstacle in the x and y directions are vx2 and vy2, respectively. When the velocities of the target obstacle and the candidate obstacle in the x and y directions are not zero, the occlusion angle ang_rad is calculated as follows:

[0083] v1 = √(vx1*vx1 + vy1*vy1)

[0084] v2 = √(vx2*vx2 + vy2*vy2)

[0085] ang_rad=cos-1(vx1 / v1*vx2 / v2+vy1 / v1*vy2 / v2)

[0086] If ang_rad is greater than π / 2, then ang_rad = π - ang_rad

[0087] The velocity feature probability trajectory_prob of the target obstacle is calculated as follows:

[0088] trajectory_prob=ang_rad / π / 2

[0089] Optionally, when the velocity of the candidate obstacle in the x and y directions is not 0 and the velocity of the target obstacle in the x and y directions is 0, the occlusion radian angle of the target obstacle is determined based on the velocities of the target obstacle and the candidate obstacle in the x and y directions, and the velocity feature probability of the target obstacle is determined based on the occlusion radian angle.

[0090] For example, the velocities of the candidate obstacle in the x and y directions are vx1 and vy1, respectively; the velocities of the target obstacle in the x and y directions are vx2 and vy2, respectively. When the velocities of the candidate obstacle in the x and y directions are not 0, and the velocities of the target obstacle in the x and y directions are 0, the occlusion angle ang_rad is calculated as follows:

[0091] ang_rad = tan-1(vy1, vx1)

[0092] If |ang_rad| is greater than π / 2, then ang_rad = π - |ang_rad|

[0093] The velocity feature probability trajectory_prob of the target obstacle is calculated as follows:

[0094] trajectory_prob=|ang_rad| / π / 2

[0095] Optionally, when the velocity of the candidate obstacle in the x and y directions is 0, the velocity of the target obstacle in the x and y directions is 0, and the acceleration of the candidate obstacle in the x and y directions is not 0, the occlusion radian angle of the target obstacle is determined based on the acceleration of the candidate obstacle in the x and y directions, and the velocity feature probability of the target obstacle is determined based on the occlusion radian angle.

[0096] For example, the accelerations of the candidate obstacle in the x and y directions are ax1 and ay1, respectively. When the velocities of the candidate obstacle and the target obstacle in the x and y directions are both 0, and the accelerations of the candidate obstacle in the x and y directions are not 0, the occlusion angle ang_rad is calculated as follows:

[0097] ang_rad = tan-1(ay1, ax1)

[0098] If |ang_rad| is greater than π / 2, then ang_rad = π - |ang_rad|

[0099] The velocity feature probability trajectory_prob of the target obstacle is calculated as follows:

[0100] trajectory_prob=|ang_rad| / π / 2

[0101] Optionally, when the velocity of the target obstacle and the candidate obstacle in the x and y directions is 0, and the acceleration of the target obstacle and the candidate obstacle in the x and y directions is 0, the velocity feature probability of the target obstacle is 0.

[0102] S330. Based on the occlusion ratio, the mirror high reflection ratio, and the velocity characteristic probability, a probability calculation is performed to determine the probability of the mirror reflection ghost image of the target obstacle.

[0103] Specifically, the occlusion ratio, specular reflection ratio, and velocity feature probability of the target obstacle are obtained, and the specular reflection ghost probability of the target obstacle is calculated based on the occlusion ratio, specular reflection ratio, and velocity feature probability of the target obstacle.

[0104] For example, the occlusion ratios covered_ratio_rad and covered_ratio_height of the target obstacle, the high intensity_ratio of the specular reflection, and the velocity feature probability trajectory_prob are obtained. The specular reflection probability can be represented by "prob", and the calculation method of the specular reflection probability prob is as follows:

[0105] prob=(covered_ratio_rad*0.7+covered_ratio_height*0.3)*

[0106] 0.8+high_intensity_ratio+trajectory_prob.

[0107] S340. If the probability of the specular reflection ghost image of the target obstacle is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular reflection ghost image.

[0108] The technical solution of this invention acquires obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle; calculates the occlusion ratio, specular reflection ratio, and velocity feature probability of the target obstacle based on the obstacle information associated with the target obstacle; performs probability calculation based on the occlusion ratio, specular reflection ratio, and velocity feature probability to determine the specular reflection ghost probability of the target obstacle, calculating the probability of determining the target obstacle as a specular reflection ghost in three dimensions, thereby improving the probability of identifying specular reflection ghosts and reducing identification errors; if the specular reflection ghost probability of the target obstacle is greater than a preset ghost probability threshold, then the target obstacle is identified as a specular reflection ghost. When the probability of the target obstacle as a specular reflection ghost meets the threshold requirement, it is identified as a specular reflection ghost, reducing the error rate of obstacle identification and further improving the accuracy of specular reflection ghosts. This achieves accurate identification of specular reflection ghosts around the target vehicle, solving the technical problem of not being able to identify lidar specular reflection ghosts in the prior art, improving the autonomous driving performance of the vehicle, and ensuring the safety of autonomous driving.

[0109] Optional, Figure 4 A flowchart illustrating another method for detecting ghost images from mirror reflections using a lidar system, provided as an embodiment of the present invention.

[0110] S410, Obstacle List. Run the obstacle detection and tracking algorithm and obtain the obstacle list obj_vec. Each obstacle contains obstacle position information center(x,y,z), obstacle size information size(w,h,l), obstacle point cloud quantity n, point cloud data ID list instance_id, and obstacle velocity information velocity v and acceleration a.

[0111] S420, Obstacle Filtering. The filtering range for specular reflection ghosting is based on the target vehicle as the coordinate origin. The lower and upper boundaries of the filtering in the x-direction are represented by "select_x_min" and "select_x_max" respectively, and the upper and lower boundaries of the filtering in the y-direction are represented by "select_y_min" and "select_y_max" respectively. The filtering range for specular reflection ghosting can be expressed as:

[0112] select_x_min <center.x<select_x_max

[0113] select_y_min <center.y<select_y_max

[0114] By filtering the obstacles around the target vehicle using the ghosting filtering range and obstacle location information through mirror reflection, multiple filtered target obstacles are obtained. These multiple target obstacles are then stored sequentially in the candidate ghosting list ghost_candidate, while maintaining the corresponding obstacle information for each target obstacle.

[0115] S430. Calculate the target information of the target obstacle. Iterate through the point cloud data ID list instance_id of each point in the target obstacle, obtain the index of each point in the obstacle, and then combine it with the original point cloud information to obtain point = pointcloud[index] of each point in the obstacle. During the traversal, update the maximum and minimum values ​​of the points in the x, y, and z directions of the target obstacle and the number of high reflection points in the obstacle, high_i_num. Obtain the four corner points (x_max, y_max), (x_max, y_min), (x_min, y_min), and (x_min, y_max) of the bounding box of the target obstacle on the xy plane. Then calculate the angle value rad of the four corner points in the polar coordinate system. By comparing the obtained four rad values, the maximum and minimum values ​​rad_min and rad_max of the obstacle in the radian direction can be obtained. The height span (height_min) and height_max of the obstacle can be directly obtained from the maximum and minimum values ​​in the z direction of the obstacle.

[0116] S440. Calculate the occlusion ratio and specular reflection ratio. Based on the radian spans (rad_min, rad_max) and height spans (height_min, height_max) of each target obstacle in the polar coordinate system, calculate the occlusion ratio between obstacles in the radian and height directions.

[0117] `covered_ratio_rad` and `covered_ratio_height`. The calculation method for the mirror high reflection ratio is as follows:

[0118] high_intensity_ratio=high_i_num / n

[0119] Specifically, the pseudocode for calculating the occlusion ratio is as follows:

[0120] input:min_back,max_back,min_ahead,max_ahead

[0121] output: covered_ratio

[0122] start

[0123] if(min_back>min_ahead&&max_back <max_ahead){

[0124] covered_ratio = 1;

[0125] }else if(min_back<min_ahead&&max_back> max_ahead){

[0126] covered_ratio=(max_ahead-min_ahead) / (max_back-min_back);

[0127] }else if(max_ahead <max_back){

[0128] covered_ratio=(max_ahead-min_back) / (max_back-min_back);

[0129] }else if(max_ahead>max_back){

[0130] covered_ratio=(max_back-min_ahead) / (max_back-min_back);

[0131] }else{

[0132] covered_ratio = -1;

[0133] }

[0134] end

[0135] The input consists of the maximum and minimum values ​​of the radii or height of the two obstacles. "back" represents the target obstacle that is farther away from the target vehicle, and "ahead" represents other obstacles that are closer to the target vehicle.

[0136] S450. Calculate the velocity feature probability. Set the velocities of the candidate obstacle in the x and y directions as vx1 and vy1; and the velocities of the target obstacle in the x and y directions as vx2 and vy2. When the velocities of both the target obstacle and the candidate obstacle in the x and y directions are not zero, the occlusion angle ang_rad is calculated as follows:

[0137] v1 = √(vx1*vx1 + vy1*vy1)

[0138] v2 = √(vx2*vx2 + vy2*vy2)

[0139] ang_rad=cos-1(vx1 / v1*vx2 / v2+vy1 / v1*vy2 / v2)

[0140] If ang_rad is greater than π / 2, then ang_rad = π - ang_rad

[0141] The velocity feature probability trajectory_prob of the target obstacle is calculated as follows:

[0142] trajectory_prob=ang_rad / π / 2

[0143] The velocities of the candidate obstacle in the x and y directions are vx1 and vy1, respectively; the velocities of the target obstacle in the x and y directions are vx2 and vy2, respectively. When the velocities of the candidate obstacle in the x and y directions are not zero, and the velocities of the target obstacle in the x and y directions are zero, the occlusion angle ang_rad is calculated as follows:

[0144] ang_rad = tan-1(vy1, vx1)

[0145] If |ang_rad| is greater than π / 2, then ang_rad = π - |ang_rad|

[0146] The velocity feature probability trajectory_prob of the target obstacle is calculated as follows:

[0147] trajectory_prob=|ang_rad| / π / 2

[0148] The accelerations of the candidate obstacle in the x and y directions are set as ax1 and ay1, respectively. When the velocities of the candidate obstacle and the target obstacle in the x and y directions are both 0, and the accelerations of the candidate obstacle in the x and y directions are not 0, the occlusion angle ang_rad is calculated as follows:

[0149] ang_rad = tan-1(ay1, ax1)

[0150] If |ang_rad| is greater than π / 2, then ang_rad = π - |ang_rad|

[0151] The velocity feature probability trajectory_prob of the target obstacle is calculated as follows:

[0152] trajectory_prob=|ang_rad| / π / 2

[0153] When the velocity of the target obstacle and the candidate obstacle in the x and y directions is 0, and the acceleration of the target obstacle and the candidate obstacle in the x and y directions is 0, the velocity feature probability of the target obstacle is 0.

[0154] S460. Calculate the specular reflection ghost probability. Obtain the occlusion ratios covered_ratio_rad and covered_ratio_height of the target obstacle, the high intensity_ratio of the specular reflection, and the velocity feature probability trajectory_prob. The specular reflection probability can be represented by "prob", and the calculation method for the specular reflection probability prob is as follows:

[0155] prob=(covered_ratio_rad*0.7+covered_ratio_height*0.3)*0.8+high_intensity_ratio+trajectory_prob.

[0156] The embodiments of the present invention achieve accurate identification of specular reflection ghost images around a target vehicle, solving the technical problem of not being able to identify specular reflection ghost images of lidar in the prior art, improving the autonomous driving performance of the vehicle, and ensuring the safety of autonomous driving.

[0157] Example 3

[0158] Figure 5 This is a schematic diagram of the structure of a laser radar specular reflection ghost detection device provided in Embodiment 3 of the present invention. Figure 5 As shown, the device includes: a lidar data transmission module 510, a specular reflection ghost detection module 520, and a specular reflection ghost judgment module 530, wherein,

[0159] The lidar data transmission module 510 is used to acquire obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle;

[0160] The specular reflection ghost detection module 520 is used to determine the specular reflection ghost probability corresponding to the target obstacle based on the obstacle information associated with the target obstacle, wherein the specular reflection ghost probability is used to represent the probability that the target obstacle is a specular reflection ghost.

[0161] The specular reflection ghost detection module 530 is used to determine the target obstacle as a specular reflection ghost if the probability of the specular reflection ghost of the target obstacle is greater than a preset ghost probability threshold.

[0162] The technical solution of this invention improves the efficiency of obstacle information collection by acquiring obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle, thereby increasing the probability of identifying specular reflection ghost images. For the target obstacle to be detected, the probability of a specular reflection ghost image corresponding to the target obstacle is determined based on the obstacle information associated with the target obstacle. The specular reflection ghost image probability represents the probability that the target obstacle is a specular reflection ghost image. For the target obstacle, the obstacle information is used to identify the target obstacle, determine the characteristics of the target obstacle, and compare... This method analyzes the probability of a target obstacle being a specular ghost image, thereby increasing the probability of identifying specular ghost images. If the probability of a target obstacle being a specular ghost image is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular ghost image. When the probability of a target obstacle being a specular ghost image meets the threshold requirement, it is identified as a specular ghost image, reducing the error rate of obstacle identification and further improving the accuracy of specular ghost image identification. This achieves accurate identification of specular ghost images around the target vehicle, solving the technical problem of not being able to identify specular ghost images in existing LiDAR technology, improving the autonomous driving performance of the vehicle, and ensuring the safety of autonomous driving.

[0163] Optionally, the specular reflection ghost detection module is specifically used for:

[0164] The occlusion ratio, specular reflection ratio, and velocity feature probability of the target obstacle are calculated based on the obstacle information associated with the target obstacle.

[0165] The probability of the specular reflection ghost image of the target obstacle is determined by performing probability calculations based on the occlusion ratio, the specular high reflectivity ratio, and the velocity characteristic probability.

[0166] Optionally, the specular reflection ghost detection module is further used for:

[0167] Based on the obstacle location information and the obstacle size information, determine the occlusion ratio between the target obstacle and other obstacles;

[0168] The mirror reflection ratio of the target obstacle is determined based on the obstacle point cloud index information;

[0169] Based on the obstacle speed information and the obstacle distance information, the speed characteristic probability of the target obstacle is determined.

[0170] Optionally, the specular reflection ghost detection module is further used for:

[0171] Calculate the distance relationship between the target obstacle and other obstacles based on the obstacle location information;

[0172] Based on the distance relationship and the obstacle size information, the occlusion ratio between the target obstacle and other obstacles is determined.

[0173] Optionally, the specular reflection ghost detection module is further used for:

[0174] Acquire the raw point cloud information detected by the lidar;

[0175] The number of specular reflections and the number of obstacle point clouds of the target obstacle are determined based on the original point cloud information, the obstacle point cloud index information, and the obstacle size information.

[0176] The mirror reflection ratio of the target obstacle is determined based on the number of mirror reflections and the number of obstacle point clouds.

[0177] Optionally, the specular reflection ghost detection module is further used for:

[0178] Based on the obstacle distance information, candidate obstacles for the target obstacle are determined;

[0179] The occlusion radian angle of the target obstacle is calculated based on the obstacle velocity information of the target obstacle and the obstacle velocity information of the candidate obstacles;

[0180] The velocity characteristic probability of the target obstacle is determined based on the occlusion radian angle.

[0181] The device further includes: a ghost range acquisition module and an obstacle filtering module; wherein...

[0182] The ghosting range acquisition module is used to acquire the specular reflection ghosting filtering range of the target vehicle;

[0183] The obstacle filtering module is used to determine the target obstacle of the target vehicle based on the obstacle location information and the specular reflection ghosting filtering range.

[0184] The laser radar specular reflection ghost detection device provided in this embodiment of the invention can execute the laser radar specular reflection ghost detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0185] Example 4

[0186] Figure 6A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0187] like Figure 6 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0188] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0189] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the method for detecting ghosting reflections on lidar surfaces.

[0190] In some embodiments, the method for detecting ghosting of lidar specular reflections can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the lidar specular reflection ghosting detection method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the lidar specular reflection ghosting detection method by any other suitable means (e.g., by means of firmware).

[0191] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0192] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0193] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0194] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0195] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0196] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0197] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0198] Example 5

[0199] This embodiment provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the program implements the steps of the laser radar specular reflection ghost detection method provided in any embodiment of the present invention. The method includes:

[0200] Obtain obstacle information corresponding to multiple target obstacles collected by the lidar in the target vehicle;

[0201] For the target obstacle to be detected, the probability of a specular reflection ghost image corresponding to the target obstacle is determined based on the obstacle information associated with the target obstacle, wherein the specular reflection ghost image probability is used to represent the probability that the target obstacle is a specular reflection ghost image;

[0202] If the probability of a specular reflection ghost image of the target obstacle is greater than a preset ghost image probability threshold, then the target obstacle is identified as a specular reflection ghost image.

[0203] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. For example, a computer-readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0204] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0205] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0206] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0207] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computing device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0208] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0209] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for detecting a mirror reflection ghost of a lidar, characterized in that, The method comprises: obtaining obstacle information corresponding to a plurality of target obstacles collected by a laser radar in a target vehicle; for a target obstacle to be detected, determining a mirror reflection ghost probability corresponding to the target obstacle according to the obstacle information associated with the target obstacle, wherein the mirror reflection ghost probability represents a probability that the target obstacle is a mirror reflection ghost; if the mirror reflection ghost probability of the target obstacle is greater than a preset ghost probability threshold, determining the target obstacle as a mirror reflection ghost; the determination of the mirror reflection ghost probability of the target obstacle according to the obstacle information comprises: calculating a shielding ratio, a mirror reflection ratio and a speed feature probability of the target obstacle according to the obstacle information associated with the target obstacle; performing probability calculation according to the shielding ratio, the mirror reflection ratio and the speed feature probability to determine the mirror reflection ghost probability of the target obstacle; the obstacle information comprises obstacle position information, obstacle point cloud index information, obstacle size information and obstacle speed information; the calculation of the shielding ratio, the mirror reflection ratio and the speed feature probability of the target obstacle according to the obstacle information comprises: determining a shielding ratio of the target obstacle to other obstacles according to the obstacle position information and the obstacle size information; determining a mirror reflection ratio of the target obstacle according to the obstacle point cloud index information; determining a speed feature probability of the target obstacle according to the obstacle speed information and the obstacle distance information; the determination of the speed feature probability of the target obstacle according to the obstacle speed information and the obstacle distance information comprises: determining a candidate obstacle of the target obstacle according to the obstacle distance information; calculating a shielding radian angle of the target obstacle according to the obstacle speed information of the target obstacle and the obstacle speed information of the candidate obstacle; wherein the shielding radian angle is an radian angle of shielding between a motion trajectory of the target obstacle and a motion trajectory of the candidate obstacle; determining the speed feature probability of the target obstacle according to the shielding radian angle.

2. The method of claim 1, wherein, the determination of the shielding ratio of the target obstacle to other obstacles according to the obstacle position information and the obstacle size information comprises: calculating a distance relationship of the target obstacle to other obstacles according to the obstacle position information; determining a shielding ratio of the target obstacle to other obstacles according to the distance relationship and the obstacle size information.

3. The method of claim 1, wherein, the determination of the mirror reflection ratio of the target obstacle according to the obstacle point cloud index information and the obstacle size information comprises: obtaining original point cloud information detected by the laser radar; determining a mirror reflection quantity of the target obstacle and an obstacle point cloud quantity according to the original point cloud information, the obstacle point cloud index information and the obstacle size information; determining the mirror reflection ratio of the target obstacle according to the mirror reflection quantity and the obstacle point cloud quantity.

4. The method of claim 1, wherein, in the obtaining of the obstacle information corresponding to the plurality of target obstacles collected by the laser radar in the target vehicle, further comprising: acquire a specular ghost filtering range of the target vehicle; determine a target obstacle of the target vehicle according to the obstacle position information and the specular ghost filtering range.

5. A device for detecting a mirror reflection ghost of a lidar, characterized in that, comprise: a laser radar data transmission module, configured to acquire obstacle information detected by a laser radar in a target vehicle; a specular ghost detection module, configured to determine a specular ghost probability of a target obstacle according to the obstacle information; a specular ghost judgment module, configured to determine the target obstacle as a specular ghost if the specular ghost probability of the target obstacle is greater than a preset ghost threshold. The specular ghost detection module is specifically configured to: calculate a shielding ratio, a specular high reflection ratio, and a speed feature probability of the target obstacle according to the obstacle information associated with the target obstacle; determine the specular ghost probability of the target obstacle according to the shielding ratio, the specular high reflection ratio, and the speed feature probability through probability calculation; The specular ghost detection module is specifically further configured to: determine a shielding ratio of the target obstacle and other obstacles according to the obstacle position information and the obstacle size information; determine a specular high reflection ratio of the target obstacle according to the obstacle point cloud index information; determine a speed feature probability of the target obstacle according to the obstacle speed information and the obstacle distance information; The specular ghost detection module is specifically further configured to: determine a candidate obstacle of the target obstacle according to the obstacle distance information; calculate a shielding radian angle of the target obstacle according to the obstacle speed information of the target obstacle and the obstacle speed information of the candidate obstacle; wherein the shielding radian angle is an radian angle of shielding between a motion trajectory of the target obstacle and a motion trajectory of the candidate obstacle; determine the speed feature probability of the target obstacle according to the shielding radian angle.

6. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the laser radar specular ghost detection method of any one of claims 1-4.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions are used to enable the processor to implement the laser radar specular ghost detection method of any one of claims 1-4 when executed.

Images