Speed and distance measurement optimization method and device and electronic equipment

Abstract

The embodiment of the invention provides a speed and distance measurement optimization method and device and electronic equipment, and the method comprises the steps: collecting the object association information of a target object in a visual field range based on a sensing module in the driving process of a target vehicle; according to the object association information, to-be-used observation position information of the target object is determined; taking the to-be-used observation position information as a target observation position when the to-be-used observation position information and the object 3D position information meet a first condition and the 2D ranging determined based on the bounding box information meets a second condition; further, determining a 2D observation speed, a 3D observation speed and an angular point observation speed corresponding to the current frame; and determining the target observation speed of the target object according to the observation speed, and determining the target speed of the target object based on the target observation speed. According to the technical scheme, the dependence on the generalization ability of the monocular 3D model is reduced, the influence of a special scene on target speed measurement and distance measurement is reduced, and the stability and accuracy of speed measurement and distance measurement are improved.

Description

Technical Field

[0001] This invention relates to the fields of automotive safety and intelligent driving technology, and in particular to an optimized method, apparatus, and electronic device for speed and distance measurement. Background Technology

[0002] Target speed and distance measurement is an essential step in driver assistance technology. It is crucial to obtain the position and speed of targets around the vehicle for trajectory prediction and planning control. Ensuring the stability and accuracy of visual target speed and distance measurement is of great significance and value.

[0003] Currently, existing methods mainly rely on monocular 3D (Mono3D) monitoring to detect the target's 3D longitudinal position, orientation, size, and 2D bounding box information. The target's position and velocity are then obtained through vehicle positioning information, Kalman filtering, motion compensation, and other methods.

[0004] However, when the vehicle is in special scenarios such as tunnels, uphill and downhill sections, or at long distances, it cannot adapt to the complex changes brought about by these special scenarios. This results in inaccurate 3D longitudinal position of the perceived target, abnormal target speed measurement, and slow convergence, which in turn causes downstream adaptive cruise control (ACC) to brake erroneously. Summary of the Invention

[0005] This invention provides an optimized method, apparatus, and electronic device for speed and distance measurement, which reduces the impact of special scenarios on target speed and distance measurement and improves the stability and accuracy of speed and distance measurement.

[0006] In a first aspect, embodiments of the present invention provide an optimized method for speed and distance measurement, comprising:

[0007] During the driving of the target vehicle, based on the perception module deployed in the target vehicle, object association information of at least one target object within the field of view is collected. The object association information includes the target object's 3D position information, object size information, object orientation information, and bounding box information including the target object.

[0008] For each object association information, the target object's observation position information in three-dimensional space is determined based on the bounding box information, the target camera's focal length, the camera's optical center information, the object's size information, and the object's orientation information; among which, the target camera is associated with the perception module.

[0009] When the observation location information to be used and the 3D location information of the object satisfy the first condition, and the 2D ranging determined based on the bounding box information satisfies the second condition, the observation location information to be used is used as the target observation location.

[0010] Based on the target observation position and the historical observation position corresponding to the previous frame, determine the 2D observation speed corresponding to the current frame; based on the object's 3D position information and the historical position information of the previous frame, determine the 3D observation speed of the current frame; and based on the corner point 3D position information of each of the eight corner points corresponding to the target object and the historical corner point position of the previous frame, determine the corner point observation speed of each corner point in the current frame.

[0011] Based on the 2D observation speed, 3D observation speed, and corner observation speed, the target observation speed of the target object is determined, and the target speed of the target object is determined based on the target observation speed.

[0012] Secondly, embodiments of the present invention also provide an optimization device for speed and distance measurement, applied to a target vehicle, the optimization device comprising:

[0013] The object association information acquisition module is used to acquire object association information of at least one target object within the field of view based on the perception module deployed in the target vehicle during the driving process of the target vehicle. The object association information includes the target object's 3D position information, object size information, object orientation information, and bounding box information including the target object.

[0014] The module for determining the observation location information to be used is used to determine the observation location information of the target object in three-dimensional space for each object association information based on the bounding box information, the camera focal length of the target camera, the camera optical center information, the object size information, and the object orientation information in the object association information; wherein, the target camera is related to the perception module;

[0015] The condition judgment module is used to use the observation location information to be used as the target observation location when the observation location information to be used and the 3D location information of the object meet the first condition, and the 2D ranging determined based on the bounding box information meets the second condition.

[0016] The observation speed determination module is used to determine the 2D observation speed corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame, and to determine the 3D observation speed of the current frame based on the object's 3D position information and the historical position information of the previous frame, and to determine the corner observation speed of each corner point in the current frame based on the corner 3D position information of each of the eight corner points corresponding to the target object and the historical corner point position of the previous frame.

[0017] The target velocity determination module is used to determine the target observation velocity of the target object based on the 2D observation velocity, 3D observation velocity, and corner observation velocity, so as to determine the target velocity of the target object based on the target observation velocity.

[0018] Thirdly, embodiments of the present invention also provide an electronic device, comprising:

[0019] At least one processor; and

[0020] A memory that is communicatively connected to at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor can perform an optimized method for speed and distance measurement as provided in any embodiment of the present invention.

[0022] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions for causing a processor to execute an optimized method for speed and distance measurement as provided in any embodiment of the present invention.

[0023] This invention, in its embodiments, collects object association information of at least one target object within the field of view of a target vehicle during its movement, based on a perception module deployed in the vehicle. For each object association information, the invention determines the target object's intended observation position information in three-dimensional space based on the bounding box information, the target camera's focal length, the camera's optical center information, the object's size information, and the object's orientation information. When the intended observation position information and the object's 3D position information satisfy a first condition, and the 2D ranging determined based on the bounding box information satisfies a second condition, the intended observation position information is used as the target observation position. The invention then determines the 2D observation speed corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame; the 3D observation speed of the current frame is determined based on the object's 3D position information and the historical position information of the previous frame; and the corner observation speed of each of the eight corner points corresponding to the target object is determined based on the 3D corner position information of each corner point and the historical corner point position of the previous frame. Finally, the invention determines the target observation speed of the target object based on the 2D observation speed, the 3D observation speed, and the corner observation speed. It reduces the reliance on the generalization ability of monocular 3D models, reduces the impact of special scenarios on target velocity and distance measurement, and improves the stability and accuracy of velocity and distance measurement.

[0024] 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

[0025] 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.

[0026] Figure 1 A flowchart illustrating an optimized method for speed and distance measurement provided in an embodiment of the present invention;

[0027] Figure 2 A flowchart illustrating an optimized method for speed and distance measurement provided in an embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of a target observation location determination method provided in an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram illustrating how to obtain the longitudinal position of a target center point according to an embodiment of the present invention.

[0030] Figure 5 This is a schematic diagram illustrating how to obtain the lateral position of a target center point according to an embodiment of the present invention.

[0031] Figure 6 This is an overall framework diagram of an optimized method for speed and distance measurement provided in an embodiment of the present invention;

[0032] Figure 7 A schematic diagram of the structure of an optimized speed and distance measuring device provided in an embodiment of the present invention;

[0033] Figure 8 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. Detailed Implementation

[0034] 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.

[0035] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0036] Figure 1 This is a flowchart illustrating an optimized speed and distance measurement method according to an embodiment of the present invention. This embodiment is applicable to situations involving speed and distance measurement of targets around a vehicle. The method can be executed by an optimized speed and distance measurement device, which can be implemented in hardware and / or software. This optimized speed and distance measurement device can be configured in a computing device. Figure 1 As shown, when applied to a target vehicle, the method includes:

[0037] S110. During the driving of the target vehicle, based on the perception module deployed in the target vehicle, collect object association information of at least one target object within the field of view.

[0038] The object association information includes the target object's 3D position information, object size information, object orientation information, and bounding box information.

[0039] In this embodiment, the perception module is a module in the target vehicle system used to perceive information about the surrounding environment. In this embodiment, sensors in the perception module collect relevant information about target objects within the target vehicle's field of view. Target objects include, but are not limited to, vehicles, pedestrians, and cyclists. The field of view can be understood as the maximum area that all sensors configured on the target vehicle can collect. Object association information includes information about the target object detected by the sensors, such as the object's 3D position information, object size information, object orientation information, and bounding box information including the target object. The object's 3D position information is the position information in the target's three-dimensional space obtained after perceiving the target object in the scene surrounding the target vehicle using a monocular 3D perception algorithm, including lateral distance, longitudinal distance, and height. The object size information represents the geometric size of the target object, including but not limited to the length, width, and height of the target object itself. The object orientation information can be understood as the facing direction of the target object relative to the target vehicle. The bounding box information including the target object is a regular geometric box that encloses the target object; the shape of the bounding box can be rectangular. This bounding box information includes, but is not limited to, the coordinates of the upper left and lower right corners of the bounding box.

[0040] For example, based on the vehicle coordinate system, if the object is parallel to the target vehicle and facing forward, the object orientation information is 0°; if it is facing left relative to the target vehicle, the object orientation information is -180°; if it is facing right relative to the target vehicle, the object orientation information is 180°.

[0041] Specifically, during the driving process of the target vehicle, the system's perception module monitors the target objects in the surrounding environment of the target vehicle in real time and obtains the object association information of all target objects within the field of view.

[0042] Optionally, the target object includes at least a vehicle, and the vehicle is in motion.

[0043] In this embodiment, the target object focuses on vehicles in motion, but also includes pedestrians, cyclists, etc.

[0044] S120. For each object association information, based on the bounding box information, the camera focal length of the target camera, the camera optical center information, the object size information, and the object orientation information in the object association information, determine the observation position information to be used for the target object in three-dimensional space.

[0045] The target camera is associated with the perception module.

[0046] In this embodiment, the target camera is a sensor deployed in the perception module of the target vehicle system, used to acquire images of the scene surrounding the target vehicle in real time and transmit the image data to the perception module. The camera focal length is the distance from the optical center of the camera lens to the imaging plane; as one of the core parameters of the camera, it determines the lens's imaging range and the size of the object's image on the imaging plane. In this embodiment, it is mainly used to calculate the target's position in the camera coordinate system. The camera optical center information is the intersection of the camera's optical coordinate system and the image coordinate system, corresponding to a specific pixel on the imaging plane, usually represented by pixel coordinates. The observation position information to be used is the three-dimensional position information calculated by combining the observation information acquired by the target camera and converted from the camera coordinate system to the vehicle coordinate system, used for subsequent determination of the target object's speed and distance.

[0047] In this embodiment, after obtaining the bounding box information, the width, height, and center point coordinates of the bounding box need to be calculated based on the vertex coordinates for subsequent calculations. For example, the bounding box information obtained from the object association information includes the coordinates of the top-left vertex (l, t) and the bottom-right vertex (r, b). Based on the two vertex coordinates of the bounding box, the width w, height h, and center point coordinates (bx, by) of the bounding box are calculated:

[0048]

[0049]

[0050]

[0051]

[0052] Specifically, the target camera captures real-time images of the scene surrounding the target vehicle and collects object association information for all target objects in the scene. Using the object size, orientation, and bounding box information from the object association information, and based on the camera's focal length and optical center information, the target object's position in three-dimensional space is calculated for observation.

[0053] S130. When the observation location information to be used and the 3D location information of the object satisfy the first condition, and the 2D ranging determined based on the bounding box information satisfies the second condition, the observation location information to be used is taken as the target observation location.

[0054] The first condition is a rule for determining the difference between the observation location information to be used and the 3D position information of the object. This boundary value is used to measure the prediction error between the observation location information to be used, derived from 2D camera perception, and the 3D position information of the object directly output by a monocular 3D algorithm. 2D ranging is the relative distance between the target vehicle and the target object calculated based on the observation location information to be used. The second condition is a rule for determining the relative distance between the target vehicle and the target object. The target observation position is the finally determined position information of the target object relative to the target vehicle.

[0055] Specifically, the system acquires the observation location information derived from 2D camera perception and calculates the difference between it and the 3D location information directly obtained by the monocular 3D algorithm. The difference is then used to determine if it meets the first condition. A 2D distance is measured from the observation location information, and this distance is used to determine if it meets the second condition. If both conditions are met, the observation location information is used as the target observation location.

[0056] S140. Determine the 2D observation velocity corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame. Also, determine the 3D observation velocity of the current frame based on the object's 3D position information and the historical position information of the previous frame. Finally, determine the corner observation velocity of each corner point in the current frame based on the corner point 3D position information of each of the eight corner points corresponding to the target object and the historical corner point position of the previous frame.

[0057] In this context, the historical observation position refers to the observation position obtained at a historical moment using different ranging methods. In this embodiment, the observation position corresponding to the previous frame at the current moment is calculated using 2D, 3D, and corner ranging methods. The 2D observation speed is the observation speed calculated using the target observation position obtained by the 2D ranging method and the historical observation position. The 3D observation speed is the observation speed calculated using the object's 3D position information output by the monocular 3D perception algorithm and the historical position information of the previous frame. The eight corner points corresponding to the target object are eight vertices derived from the target object association information obtained by the system using the monocular 3D perception algorithm. The corner point 3D position information is the position information of the eight corner points in the vehicle coordinate system. The corner point observation speed is the target observation speed calculated using the corner point 3D position information and the corner point position of the previous frame.

[0058] Specifically, the system uses a 2D camera, 3D perception algorithms, and corner ranging methods to calculate the current observation position of the target object and stores it in the module for velocity measurement in the next moment. When calculating the observed velocity of the target object, the system obtains the target observation position corresponding to the current frame and the historical observation positions from the previous frame to calculate the 2D observed velocity of the current frame. Similarly, the system obtains the 3D position of the object in the current frame and the 3D corner positions of the eight corner points of the target object, and combines this information with the corresponding historical positions from the previous frame to calculate the 3D observed velocity and corner observed velocity of the target object.

[0059] Optionally, before determining the corner observation velocity of each corner point in the current frame based on the 3D position information of each of the eight corner points corresponding to the target object and the historical corner point positions of the previous frame, the method further includes:

[0060] Based on the center point location and size information of the target object, determine the 3D position information of the eight corner points associated with the target object.

[0061] The center point position is the coordinate of the center point in the 3D position information of the object output by the monocular 3D algorithm.

[0062] Specifically, obtain the center point coordinates and object size information of the target object, such as its length, width, and height. Based on this information, calculate the 3D coordinates of the eight corner points of the target object.

[0063] Optionally, the 2D observation velocity corresponding to the current frame is determined based on the target observation location and the historical observation location corresponding to the previous frame, including:

[0064] The 2D observation velocity of the current frame is determined based on the target observation position, the historical observation position corresponding to the previous frame, and the interval between two frames. Correspondingly, the 3D observation velocity of the current frame is determined based on the object's 3D position information and the historical position information of the previous frame, including: determining the 3D observation velocity of the current frame based on the object's 3D position information, the historical position information of the previous frame, and the interval between two adjacent frames. Correspondingly, the corner observation velocity of each corner point in the current frame is determined based on the corner point 3D position information of each of the eight corner points corresponding to the target object and the historical corner point position of the previous frame, including: for each corner point, determining the corner observation velocity of the corner point based on the corner point 3D position of the corner point and the historical corner point position of the corner point in the previous frame.

[0065] The interval between two frames is the time difference between the previous frame and the current time. For example, if the acquired target observation position is pos... obs The historical observation position of the previous frame is pos last And the calculated interval between the two frames is Then calculate the observation velocity of the current frame:

[0066]

[0067] Specifically, the target observation position, 3D position information, and corner 3D position information are obtained using 2D, 3D, and corner ranging methods, respectively. The difference between these two information and the corresponding historical observation position from the previous frame is calculated and divided by the time difference between the two frames. Using this method, the 2D observation velocity, 3D observation velocity, and corner observation velocity of each corner point in the current frame are obtained.

[0068] S150. Based on the 2D observation speed, 3D observation speed and corner observation speed, determine the target observation speed of the target object, and determine the target speed of the target object based on the target observation speed.

[0069] The target observation velocity is the final observation velocity obtained by comparing the 2D observation velocity, 3D observation velocity, and corner velocity. The target velocity is the final velocity information of the target object output after fusing the target observation velocity of the current frame with the target velocity of the previous frame through Kalman filtering.

[0070] Specifically, based on the calculated 2D observation velocity, 3D observation velocity, and corner observation velocity, the minimum value among these three is selected as the target observation velocity for the current frame. The filtered target observation velocity from the previous frame is then input into the Kalman filter model along with the target observation velocity of the current frame. A weighted fusion of the target velocities from the previous and current frames is performed using a filtering algorithm, ultimately outputting the final velocity of the target object in the current frame, i.e., the target velocity.

[0071] Optionally, the target observation speed of the target object is determined based on the 2D observation speed, 3D observation speed, and corner observation speed, including:

[0072] Based on the 2D observation velocity, determine the first average velocity of the target object within a preset number of historical frames; based on the 3D observation velocity, determine the second average velocity of the target object within a preset number of historical frames; based on the minimum value among the corner observation velocities, determine the third average velocity of the target object within a preset number of historical frames; and take the minimum value among the first, second, and third average velocities as the target observation velocity.

[0073] The historical preset frame count is a pre-set number of frames used to calculate the average velocity. The first average velocity is the average velocity of the target object determined using the 2D ranging method within the historical preset frame count; the second average velocity is the average velocity of the target object determined using the monocular 3D perception algorithm within the historical preset frame count; and the third average velocity is the average velocity of the target object determined using the corner ranging method within the historical preset frame count.

[0074] Specifically, a historical preset frame number is set, and the target prediction position of the historical preset frame number determined using the 2D ranging method, the 3D position information of the historical preset frame number determined using the monocular 3D perception algorithm, and the corner 3D position information of the eight corner points of the historical preset frame number determined using the corner ranging method are obtained. The first average velocity, second average velocity, and third average velocity are obtained according to the average velocity calculation formula. The minimum value among the first average velocity, second average velocity, and third average velocity is selected as the target observation velocity.

[0075] For example, if the historical preset frame number is set to 10, the target prediction position vel of the historical preset frame number is first obtained through the 2D ranging method. frame Then, the predicted locations of the above 10 targets are summed, and finally the sum is divided by the historical preset number of frames to obtain the first average velocity vel. 2D :

[0076]

[0077] Acquire 3D position information of 10 objects determined using 3D perception algorithms. frame The 3D position information of the above 10 objects is summed, and the sum is divided by the historical preset number of frames to obtain the second average velocity vel. 3D :

[0078]

[0079] Obtain the 3D position information of the eight corner points determined by the 3D perception algorithm. frame The 3D position information of the object in 10 frames for each corner point is summed, and the sum is divided by the preset number of historical frames to obtain the third average velocity vel of that corner point. corner=i :

[0080]

[0081] The third average velocity at the eight corner points was calculated, and the minimum value among these eight corner points was taken as the final third average velocity.

[0082]

[0083] The technical solution provided in this invention involves collecting object association information of at least one target object within the field of view of a target vehicle during its operation, based on a perception module deployed in the vehicle. For each object association information, the target object's observation position information in three-dimensional space is determined based on the bounding box information, the target camera's focal length, the camera's optical center information, the object's size information, and the object's orientation information. The observation position information is compared with the object's 3D position information to determine if a first condition is met, and the 2D ranging determined by the bounding box information is judged to meet a second condition. If both the first and second conditions are met, the observation position information is used as the target observation position. The 2D observation velocity corresponding to the current frame is determined based on the target observation position and the historical observation position corresponding to the previous frame. The 3D observation velocity of the current frame is determined based on the object's 3D position information and the historical position information of the previous frame. Finally, the corner observation velocity of each corner point in the current frame is determined based on the 3D position information of each of the eight corner points corresponding to the target object and the historical corner point positions of the previous frame. Finally, based on the 2D observation speed, 3D observation speed, and corner observation speed, the target observation speed of the target object is determined, thereby determining the target speed of the target object. This reduces the reliance on the generalization ability of the monocular 3D model, minimizes the impact of special scenes on target velocity and distance measurement, and improves the stability and accuracy of velocity and distance measurement.

[0084] Figure 2 This is a flowchart of an optimized method for speed and distance measurement provided by an embodiment of the present invention. This embodiment further refines the determination of the target observation position based on the above embodiments. Figure 2 As shown, the method includes:

[0085] S210. During the driving of the target vehicle, based on the perception module deployed in the target vehicle, collect object association information of at least one target object within the field of view.

[0086] The object association information includes the target object's 3D position information, object size information, object orientation information, and bounding box information.

[0087] S220. Based on the first height in the object size information, the second height in the bounding box information, and the camera focal length, determine the first three-dimensional longitudinal position of the target tail of the target object in the camera coordinate system.

[0088] The first height refers to the height of the target object obtained using a 3D monocular perception algorithm, and the second height refers to the height information of the bounding box. The target tail is the part of the target object closest to the target vehicle from a top-down view. The first three-dimensional longitudinal position is the longitudinal position coordinate of the target tail in the camera's three-dimensional coordinate system.

[0089] Specifically, the object association information of the target object is obtained. Based on the first height in the object size information, the second height in the bounding box information, and the camera focal length, the first three-dimensional longitudinal position of the target object's tail in the target coordinate system is calculated using the pinhole imaging principle.

[0090] For example, Figure 3 This is a schematic diagram of a target observation location determination method provided by an embodiment of the present invention. Figure 3 As shown, if the first height obtained is H, the camera focal length is focal_length, and the second height h in the bounding box information is used, the longitudinal position pos(2) of the target tail in the camera coordinate system is calculated using the pinhole imaging principle:

[0091]

[0092] S230. Using the first three-dimensional longitudinal position, the first width and first height in the object size information, and the object orientation information, determine the longitudinal position of the center point of the target object in three-dimensional space.

[0093] Here, the first width is the width of the target object in the object size information, the center point of the target object is the geometric center of the target object in three-dimensional space, and this center point is a three-dimensional coordinate point containing x, y, and z axis coordinates. The longitudinal position of the center point is the position coordinate of the center point of the target object along the longitudinal direction in the camera's three-dimensional coordinate system.

[0094] Specifically, the first three-dimensional longitudinal position, first width, first height, and object orientation information of the target object are obtained. The longitudinal position of the center point of the target object in three-dimensional space is calculated, and different calculation methods are set according to different object orientation information.

[0095] For example, Figure 4 This is a schematic diagram illustrating how to obtain the longitudinal position of a target center point according to an embodiment of the present invention. Figure 4 As shown, the longitudinal position pos_tmp(2) of the target center point in the camera coordinate system is obtained according to the different orientations of the target:

[0096]

[0097] Where pos(2) is the longitudinal position of the target tail, W is the first width, and L is the first height. This refers to the orientation information of the target object.

[0098] S240. Based on bounding box information, camera optical center information, longitudinal position of center point, and camera focal length, determine the lateral position of the center point of the target object in three-dimensional space.

[0099] Here, the horizontal position of the center point is the horizontal position coordinate of the center point of the target object in the camera's three-dimensional coordinate system.

[0100] Specifically, the bounding box information, camera optical center information, vertical position of the center point, and camera focal length are obtained from the object association information, and the horizontal position of the center point of the target object in three-dimensional space is calculated.

[0101] For example, Figure 5 This is a schematic diagram illustrating how to obtain the lateral position of a target center point according to an embodiment of the present invention. Figure 5 As shown, the center point coordinates of the target object are derived from the bounding box information. Using the center point coordinates (bx,by), camera optical center information cx_, vertical position of the center point pos_tmp(2), and camera focal length focal_length, the horizontal position of the center point of the target object in three-dimensional space is obtained:

[0102]

[0103] S250. Based on the longitudinal and lateral positions of the center point, determine the observation location information to be used.

[0104] Among them, the observation location information to be used is the observation location information of the target object in three-dimensional space derived using the 2D ranging method.

[0105] Specifically, the observation location information to be used includes the longitudinal position and the lateral position of the center point.

[0106] S260. Based on the spatial transformation matrix, perform coordinate system transformation on the horizontal and vertical positions of the center point in the observation position information to be used, so as to obtain the observation position to be used in the vehicle coordinate system.

[0107] The observation location to be used is determined based on two-dimensional bounding box information.

[0108] In this embodiment, the spatial transformation matrix is ​​a method of rotating and translating one coordinate system to another, used to characterize the relative positional relationship between the two three-dimensional coordinate systems. The vehicle coordinate system is a three-dimensional coordinate system established with the target vehicle itself as the reference center, such as using the rear axle center point of the target vehicle as the reference center.

[0109] Specifically, the calculated observation position information in the camera coordinate system is transformed into the observation position in the vehicle coordinate system with the rear axle center of the target vehicle as the reference center.

[0110] For example, the calculated observation position information pos_tmp in the camera coordinate system is obtained, and the observation position pos_2d in the vehicle coordinate system is obtained using the spatial transformation matrix:

[0111]

[0112] Here, cam2ego is the transformation matrix from the camera coordinate system to the vehicle coordinate system.

[0113] S270. Determine the longitudinal position difference based on the observation position information to be used and the 3D position information of the object.

[0114] Among them, the longitudinal position difference is the prediction error between the observation position information to be used, which is derived from 2D camera perception, and the object 3D position information directly output by the monocular 3D algorithm.

[0115] Specifically, the system obtains the coordinate system transformation information of the observation position to be used and the 3D position information of the object, and then subtracts the two to obtain the longitudinal position difference.

[0116] For example, the observation position information pos_2d(2) that has undergone coordinate system transformation is obtained, and the longitudinal position in the 3D position information of the object acquired by the monocular 3D perception algorithm is obtained. The difference between the two is used to obtain the longitudinal position difference dis_diff:

[0117]

[0118] S280. Based on the observation location to be used, determine the first distance information between the target vehicle and the target object.

[0119] The first distance information is the relative distance between the target vehicle and the target object, which is calculated from the observation position to be used in the vehicle coordinate system.

[0120] Specifically, the observation position to be used in the vehicle coordinate system is obtained, the distance from the observation position to the origin of the coordinate system is calculated, and this distance is used as the first distance information.

[0121] S290. When the longitudinal position difference is greater than or equal to a preset difference threshold, and the first distance information is greater than or equal to a preset distance threshold, the observation position to be used is taken as the target observation position.

[0122] The preset difference threshold is the boundary value used in the first pre-set condition to determine the difference between the observed location information and the 3D location information of the object. The preset distance threshold is the boundary value used in the second pre-set condition to determine the relative distance between the target vehicle and the target object.

[0123] Specifically, a preset difference threshold and a preset distance threshold are set. The preset difference threshold is compared with the longitudinal position difference, and the preset distance threshold is compared with the first distance information. If the longitudinal position difference is greater than or equal to the preset difference threshold, and the first distance information is also greater than or equal to the preset distance threshold, the observation position to be used is taken as the target prediction position.

[0124] For example, if the target object is a vehicle, the preset difference threshold is 30m, and the preset distance threshold is 80m. The calculated first distance information of the vehicle relative to the target vehicle is 80m, and the longitudinal position difference is 90m. Since the first distance information is greater than the preset distance threshold, and the longitudinal position difference is greater than the preset difference threshold, it indicates that the target object and the target vehicle are currently too far apart. The observation position to be used obtained by the 2D ranging method will be used as the target observation position.

[0125] Optionally, the method also includes:

[0126] When the longitudinal position difference is less than a preset difference threshold, or when the first distance information is less than a preset distance threshold, the object's 3D position information is used as the target observation position.

[0127] Specifically, if only the longitudinal position difference is greater than the preset difference threshold, or only the first distance information is greater than the preset distance threshold, or neither of these conditions is met, the object 3D position information directly acquired by the monocular 3D algorithm will be used as the target observation position.

[0128] The technical solution provided in this invention involves collecting object association information of at least one target object within the field of view of a target vehicle during its movement, based on a perception module deployed in the target vehicle. The first three-dimensional longitudinal position of the target object's tail in the camera coordinate system is determined based on the first height in the object's size information, the second height in the bounding box information, and the camera focal length. The longitudinal position of the target object's center point in three-dimensional space is then determined using the first three-dimensional longitudinal position, the first width, the first height in the object's size information, and the object's orientation information. Simultaneously, the lateral position of the target object's center point in three-dimensional space is determined based on the bounding box information, the camera's optical center information, the longitudinal position of the center point, and the camera focal length. Finally, the observation position information to be used is determined based on the longitudinal and lateral positions of the center point. Further, a coordinate system transformation is performed on the lateral and longitudinal positions of the center point in the observation position information to be used based on a spatial transformation matrix to obtain the observation position to be used in the vehicle coordinate system. The longitudinal position difference and the first distance information between the target vehicle and the target object are determined based on the observation position information to be used and the object's 3D position information. The system judges the difference between the first distance information and the longitudinal position. If the difference in the longitudinal position is greater than or equal to a preset difference threshold, and the first distance information is greater than or equal to a preset distance threshold, the observation position to be used is taken as the target observation position. By obtaining the observation distance and observation speed of the target object based on different ranging and velocity measurement methods, the impact of special scenarios on target velocity and ranging is reduced, improving the stability and accuracy of velocity and ranging measurements.

[0129] Figure 6 This is an overall framework diagram of an optimized speed and distance measurement method provided in an embodiment of the present invention, combined with... Figure 6 Understand the technical solutions of the embodiments of the present invention.

[0130] like Figure 6 As shown, this embodiment explains the overall implementation of the solution based on the above optional implementation methods, specifically including:

[0131] First, the system uses the Mono3D scheme to obtain the 3D position and type of the target object. Three methods are used for ranging and velocity measurement: 2D-based, 3D-based, and target corner-based methods. The 2D-based method acquires the 2D bounding box obtained using a monocular 3D algorithm and uses camera intrinsics to obtain image focal length and optical information, resulting in a 2D ranging result (observation position information to be used). The difference between the 3D and 2D longitudinal positions (longitudinal position difference) is calculated, and the relative distance between the target vehicle and the target is obtained based on the 2D ranging result (first distance information). Preset difference thresholds and preset distance thresholds are set. If the longitudinal position difference is greater than the preset difference threshold, and the first distance information is greater than the preset distance threshold, the 2D ranging result is used as the observation position. The observation velocity is calculated based on the observation position, and the average velocity of the previous ten frames (a preset number of historical frames) is calculated based on the observation position information. The 3D-based method uses the center point ranging result from the acquired 3D position information as the observation position and calculates the observation velocity accordingly. The average velocity of the previous ten frames is obtained. Corner-based method: Based on the target's 3D center point and the target object's length, width, and height, the positions of the target's corner points are obtained. The distance measurement results for the four corner points are calculated, and the observation velocity for each corner point is calculated. The minimum value is taken by comparing all the corner point observation velocities. Finally, the average velocity obtained from the three methods is compared, and the minimum value is determined as the final target observation velocity (target observation velocity).

[0132] This invention, in its embodiments, collects object association information of at least one target object within the field of view of a target vehicle during its movement, based on a perception module deployed in the vehicle. For each object association information, the invention determines the target object's intended observation position information in three-dimensional space based on the bounding box information, the target camera's focal length, the camera's optical center information, the object's size information, and the object's orientation information. When the intended observation position information and the object's 3D position information satisfy a first condition, and the 2D ranging determined based on the bounding box information satisfies a second condition, the intended observation position information is used as the target observation position. The invention then determines the 2D observation speed corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame; the 3D observation speed of the current frame is determined based on the object's 3D position information and the historical position information of the previous frame; and the corner observation speed of each of the eight corner points corresponding to the target object is determined based on the 3D corner position information of each corner point and the historical corner point position of the previous frame. Finally, the invention determines the target observation speed of the target object based on the 2D observation speed, the 3D observation speed, and the corner observation speed. It reduces the reliance on the generalization ability of monocular 3D models, reduces the impact of special scenarios on target velocity and distance measurement, and improves the stability and accuracy of velocity and distance measurement.

[0133] Figure 7 This is a schematic diagram of the structure of an optimized speed and distance measuring device provided in an embodiment of the present invention. Figure 7 As shown, the device includes: an object association information acquisition module 310, a to-be-used observation location information determination module 320, a condition judgment module 330, an observation speed determination module 340, and a target speed determination module 350.

[0134] The object association information acquisition module 310 is used to acquire object association information of at least one target object within the field of view of the target vehicle during its movement, based on a perception module deployed in the target vehicle. The object association information includes the target object's 3D position information, object size information, object orientation information, and bounding box information. The pending observation position information determination module 320 is used to determine the pending observation position information of the target object in three-dimensional space for each object association information, based on the bounding box information, the target camera's focal length, camera optical center information, object size information, and object orientation information; wherein the target camera is associated with the perception module. The condition judgment module 330 is used to use the pending observation position information as the target observation position when the pending observation position information and the object's 3D position information satisfy a first condition, and when the 2D ranging determined based on the bounding box information satisfies a second condition. The observation velocity determination module 340 is used to determine the 2D observation velocity corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame; to determine the 3D observation velocity of the current frame based on the object's 3D position information and the historical position information of the previous frame; and to determine the corner observation velocity of each corner point in the current frame based on the corner point 3D position information of each of the eight corner points corresponding to the target object and the historical corner point positions of the previous frame. The target velocity determination module 350 is used to determine the target observation velocity of the target object based on the 2D observation velocity, 3D observation velocity, and corner observation velocity, thereby determining the target velocity of the target object.

[0135] This invention, in its embodiments, collects object association information of at least one target object within the field of view of a target vehicle during its movement, based on a perception module deployed in the vehicle. For each object association information, the invention determines the target object's intended observation position information in three-dimensional space based on the bounding box information, the target camera's focal length, the camera's optical center information, the object's size information, and the object's orientation information. When the intended observation position information and the object's 3D position information satisfy a first condition, and the 2D ranging determined based on the bounding box information satisfies a second condition, the intended observation position information is used as the target observation position. The invention then determines the 2D observation speed corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame; the 3D observation speed of the current frame is determined based on the object's 3D position information and the historical position information of the previous frame; and the corner observation speed of each of the eight corner points corresponding to the target object is determined based on the 3D corner position information of each corner point and the historical corner point position of the previous frame. Finally, the invention determines the target observation speed of the target object based on the 2D observation speed, the 3D observation speed, and the corner observation speed. It reduces the reliance on the generalization ability of monocular 3D models, reduces the impact of special scenarios on target velocity and distance measurement, and improves the stability and accuracy of velocity and distance measurement.

[0136] Based on the above technical solutions, the module for determining the observation location information to be used includes:

[0137] The first three-dimensional longitudinal position determination unit is used to determine the first three-dimensional longitudinal position of the target tail of the target object in the camera coordinate system based on the first height in the object size information, the second height in the bounding box information, and the camera focal length.

[0138] The center point longitudinal position determination unit is used to determine the center point longitudinal position of the target object in three-dimensional space by using the first three-dimensional longitudinal position, the first width and the first height in the object size information, and the object orientation information.

[0139] The center point lateral position determination unit is used to determine the lateral position of the center point of the target object in three-dimensional space based on bounding box information, camera optical center information, center point longitudinal position and camera focal length.

[0140] The unit for determining the observation location information to be used is used to determine the observation location information to be used based on the longitudinal and lateral positions of the center point.

[0141] Based on the above technical solutions, the condition judgment module includes:

[0142] The observation position acquisition unit is used to perform coordinate system transformation on the horizontal and vertical positions of the center point in the observation position information to be used based on the spatial transformation matrix, so as to obtain the observation position to be used in the vehicle coordinate system. The observation position to be used is the observation position determined based on the two-dimensional bounding box information.

[0143] The longitudinal difference determination unit is used to determine the longitudinal position difference based on the observation position information to be used and the 3D position information of the object.

[0144] The first distance information determination unit is used to determine the first distance information between the target vehicle and the target object based on the observation position to be used.

[0145] The target observation location determination unit is used to determine the observation location to be used as the target observation location when the longitudinal position difference is greater than or equal to a preset difference threshold and the first distance information is greater than or equal to a preset distance threshold.

[0146] Based on the above technical solutions, the condition judgment module further includes: when the longitudinal position difference is less than a preset difference threshold, or when the first distance information is less than a preset distance threshold, the object's 3D position information is used as the target observation position.

[0147] Based on the above technical solutions, the observation velocity determination module includes:

[0148] The corner point 3D position information determination unit is used to determine the corner point 3D position information of the eight corner points associated with the target object based on the center point position and object size information of the target object.

[0149] Based on the above technical solutions, the observation velocity determination module includes:

[0150] The 2D observation velocity determination unit is used to determine the 2D observation velocity of the current frame based on the target observation position, the historical observation position corresponding to the previous frame, and the interval between the two frames.

[0151] The 3D observation speed determination unit is used to determine the 3D observation speed of the current frame based on the object's 3D position information, the historical position information of the previous frame, and the interval between two adjacent frames.

[0152] The corner observation velocity determination unit is used to determine the corner observation velocity for each corner point based on its 3D corner point position and its historical corner point position in the previous frame.

[0153] Based on the above technical solutions, the target speed determination module includes:

[0154] The first average velocity determination unit is used to determine the first average velocity of the target object within a preset number of historical frames based on the 2D observation velocity.

[0155] The second average velocity determination unit is used to determine the second average velocity of the target object within a preset number of historical frames based on the 3D observation velocity.

[0156] The third average velocity determination unit is used to determine the third average velocity of the target object within a preset number of historical frames based on the minimum value of the corner observation velocity.

[0157] The target observation velocity determination unit is used to take the minimum value among the first average velocity, the second average velocity, and the third average velocity as the target observation velocity.

[0158] Based on the above technical solutions, optionally, the target object includes at least a vehicle, and the vehicle is in motion.

[0159] The speed and distance measurement optimization device provided in this embodiment of the invention can execute the speed and distance measurement optimization method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0160] Figure 8 A schematic diagram of an electronic device 10, which 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.

[0161] like Figure 8 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 can 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.

[0162] 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.

[0163] 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 an optimized method for speed and distance measurement.

[0164] In some embodiments, an optimization of speed and distance measurement 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 speed and distance measurement optimization method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform a speed and distance measurement optimization method by any other suitable means (e.g., by means of firmware).

[0165] 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.

[0166] 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.

[0167] 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.

[0168] 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).

[0169] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or 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.

[0170] 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.

[0171] 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.

[0172] 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. An optimization method for speed and distance measurement, characterized in that, Applied to a target vehicle, the method comprises: During driving of the target vehicle, object association information of at least one target object in a field of view is collected based on a perception module deployed in the target vehicle, wherein the object association information comprises object 3D position information, object size information, object orientation information of the target object, and bounding box information comprising the target object; For each object association information, a to-be-used observation position information of the target object in a three-dimensional space is determined according to the bounding box information in the object association information, camera focal length, camera optical center information of a target camera, object size information, and the object orientation information; wherein the target camera is related to the perception module; When the to-be-used observation position information and the object 3D position information satisfy a first condition, and a 2D ranging determined based on the bounding box information satisfies a second condition, the to-be-used observation position information is taken as a target observation position; A 2D observation speed corresponding to a current frame is determined according to the target observation position and a historical observation position corresponding to a previous frame, a 3D observation speed of the current frame is determined according to the object 3D position information and historical position information of the previous frame, and an angle point observation speed of each angle point in the current frame is determined according to angle point 3D position information of each angle point in eight angle points corresponding to the target object and historical angle point positions of the previous frame; A target observation speed of the target object is determined according to the 2D observation speed, the 3D observation speed, and the angle point observation speed, so as to determine a target speed of the target object based on the target observation speed.

2. The method of claim 1, wherein, The determination of the to-be-used observation position information of the target object in the three-dimensional space according to the bounding box information in the object association information, the camera focal length, the camera optical center information of the target camera, the object size information, and the object orientation information comprises: A first three-dimensional longitudinal position of a target tail of the target object in a camera coordinate system is determined according to a first height in the object size information, a second height in the bounding box information, and the camera focal length; A center point longitudinal position of a center point of the target object in the three-dimensional space is determined according to the first three-dimensional longitudinal position, a first width in the object size information, the first height, and the object orientation information; A center point transverse position of the center point of the target object in the three-dimensional space is determined based on the bounding box information, the camera optical center information, the center point longitudinal position, and the camera focal length; The to-be-used observation position information is determined based on the center point longitudinal position and the center point transverse position.

3. The method of claim 2, wherein, The taking of the to-be-used observation position information as the target observation position when the to-be-used observation position information and the object 3D position information satisfy the first condition, and the 2D ranging determined based on the bounding box information satisfies the second condition, comprises: transform the center point horizontal position and the center point vertical position in the to-be-used observation position information into a vehicle coordinate system based on a spatial transformation matrix to obtain a to-be-used observation position in the vehicle coordinate system, wherein the to-be-used observation position is an observation position determined based on two-dimensional bounding box information; determine a vertical position difference value based on the to-be-used observation position information and the object 3D position information; determine first distance information between the target vehicle and the target object based on the to-be-used observation position; when the vertical position difference value is greater than or equal to a preset difference threshold value and the first distance information is greater than or equal to a preset distance threshold value, use the to-be-used observation position as the target observation position.

4. The method of claim 3, wherein, The method further includes: when the vertical position difference value is less than the preset difference threshold value or the first distance information is less than the preset distance threshold value, use the object 3D position information as the target observation position.

5. The method of claim 1, wherein, Before determining the corner point observation speed of each corner point in the current frame based on the corner point 3D position information of each corner point corresponding to the target object and the historical corner point position of the previous frame, the method further includes: determine the corner point 3D position information of the eight corner points associated with the target object based on the center point position of the target object and the object size information.

6. The method of claim 1, wherein, The determination of the 2D observation speed corresponding to the current frame based on the target observation position and the historical observation position corresponding to the previous frame includes: determine the 2D observation speed of the current frame based on the target observation position, the historical observation position corresponding to the previous frame, and the interval duration between the two frames; Correspondingly, the determination of the 3D observation speed of the current frame based on the object 3D position information and the historical position information of the previous frame includes: determine the 3D observation speed of the current frame based on the object 3D position information, the historical position information of the previous frame, and the interval duration between the two adjacent frames; Correspondingly, the determination of the corner point observation speed of each corner point in the current frame based on the corner point 3D position information of each corner point corresponding to the target object and the historical corner point position of the previous frame includes: for each corner point, determine the corner point observation speed of the corner point based on the corner point 3D position of the corner point and the historical corner point position of the corner point in the previous frame.

7. The method of claim 1, wherein, The determination of the target observation speed of the target object based on the 2D observation speed, the 3D observation speed, and the corner point observation speed includes: determine a first average speed of the target object within a historical preset number of frames based on the 2D observation speed; determine a second average speed of the target object within the historical preset number of frames based on the 3D observation speed; determine a third average speed of the target object within the historical preset number of frames based on the minimum value in the corner point observation speed; use the minimum value among the first average speed, the second average speed, and the third average speed as the target observation speed.

8. The method of claim 1, wherein, The target object at least includes a vehicle, and the vehicle is in a moving state.

9. An optimization device for speed and distance measurement, characterized in that The application is applied to a target vehicle and includes: An object association information collection module is configured to collect object association information of at least one target object within a field of view of the target vehicle based on a perception module deployed in the target vehicle during driving of the target vehicle, wherein the object association information includes object 3D position information, object size information, object orientation information, and bounding box information of the target object. A to-be-used observation position information determination module is configured to determine, for each object association information, to-be-used observation position information of the target object in a three-dimensional space according to the bounding box information, camera focal length, camera optical center information, object size information, and object orientation information in the object association information, wherein the target camera is associated with the perception module. A condition determination module is configured to determine the to-be-used observation position information as a target observation position when the to-be-used observation position information and the object 3D position information satisfy a first condition, and a 2D ranging determined based on the bounding box information satisfies a second condition. An observation speed determination module is configured to determine a 2D observation speed corresponding to a current frame according to the target observation position and a historical observation position corresponding to a previous frame, determine a 3D observation speed of the current frame according to the object 3D position information and historical position information of the previous frame, and determine an angle point observation speed of each angle point in the current frame according to angle point 3D position information of each angle point in eight angle points corresponding to the target object and historical angle point positions of the previous frame. A target speed determination module is configured to determine a target observation speed of the target object according to the 2D observation speed, the 3D observation speed, and the angle point observation speed, and determine a target speed of the target object based on the target observation speed.

10. An electronic device, comprising: The electronic device includes: 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 optimization method for speed measurement and ranging according to any one of claims 1-8.

Images