Method, device, equipment and storage medium for determining target position

By generating a target image and determining the three-dimensional position of the target object in combination with the second camera parameters, the problem of blind spots and incomplete detection of the object's spatial position perception in autonomous driving is solved, and the accuracy and safety of target object position detection is improved.

CN114821497BActive Publication Date: 2024-11-08GUANGZHOU WERIDE TECH LTD CO
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210171929.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-02-24
Publication Date
2024-11-08
Estimated Expiration
2042-02-24

AI Technical Summary

Technical Problem

In the field of autonomous driving, the spatial position perception of objects has problems such as blind spots in line of sight and incomplete detection of target objects, and it is difficult to accurately calculate the three-dimensional spatial distance between target objects and vehicles, resulting in safety hazards.

Method used

By acquiring the mapping matrix between at least two camera images, a target image is generated, target object detection is performed, and the three-dimensional position information of the target object is determined in combination with the second camera parameters, thereby improving the accuracy of target object position detection.

Benefits of technology

It effectively avoids blind spots in the line of sight, improves the accuracy and completeness of target object detection, and enhances the safety of autonomous driving vehicles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114821497B_ABST
    Figure CN114821497B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of image processing technology, and discloses a method, device, equipment and storage medium for determining the position of a target object, which are used to improve the accuracy of target object position detection. The method for determining the position of a target object comprises: obtaining a target mapping matrix set, the target mapping matrix set comprising mapping matrices between at least two first camera images and second camera images, one first camera image corresponding to one first camera; generating a target image according to the target mapping matrix set; performing target object detection on the target image to obtain two-dimensional detection frame information of at least one target object; obtaining second camera parameters, and determining three-dimensional position information of each target object according to the second camera parameters and the two-dimensional detection frame information of each target object.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method, device, equipment and storage medium for determining a target position. Background Art

[0002] Perceiving external objects is one of the most important capabilities in the field of autonomous driving, and clearly distinguishing various objects and determining their spatial positions is a crucial part of the perception task.

[0003] Existing technologies for perceiving the spatial position of objects usually perform target detection on images captured by a single camera based on a deep learning model. This approach often has blind spots, misses targets in some areas, and makes it difficult to accurately calculate the three-dimensional spatial distance between the target and the vehicle, posing a safety hazard to vehicle driving. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for determining the position of a target object, which are used to improve the accuracy of detecting the position of the target object.

[0005] A first aspect of the present invention provides a method for determining a position of a target object, comprising:

[0006] Acquire a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera;

[0007] generating a target image according to the target mapping matrix set;

[0008] Performing target object detection on the target image to obtain two-dimensional detection frame information of at least one target object;

[0009] The second camera parameters are obtained, and the three-dimensional position information of each target object is determined according to the second camera parameters and the two-dimensional detection frame information of each target object.

[0010] Optionally, the acquiring a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and the second camera image, respectively, including:

[0011] defining a second camera intrinsic parameter and a second camera extrinsic parameter, wherein the second camera extrinsic parameter is used to indicate a relative position parameter of the second camera to a preset reference object;

[0012] Obtaining a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, wherein the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to the second camera;

[0013] A pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is determined according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, and a mapping matrix between each first camera image and the second camera image is generated according to all pixel mapping matrices to obtain a target mapping matrix set.

[0014] Optionally, the defining of the second camera intrinsic parameter and the second camera extrinsic parameter, wherein the second camera extrinsic parameter is used to indicate a relative position parameter of the second camera to a preset reference object, includes:

[0015] Setting a second camera intrinsic parameter, where the second camera intrinsic parameter is used to indicate an image parameter of the second camera;

[0016] A second camera coordinate system and a preset reference object coordinate system are defined, and relative position parameters from the second camera to the preset reference object are calculated according to the second camera coordinate system and the preset reference object coordinate system to obtain second camera extrinsics.

[0017] Optionally, the obtaining of a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, wherein the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to the second camera, includes:

[0018] Reading a distortion intrinsic parameter corresponding to each first camera of at least two first cameras, and defining a first camera coordinate system corresponding to each first camera;

[0019] According to the first camera coordinate system and the second camera coordinate system corresponding to each first camera, the relative position parameters from each first camera to the second camera are calculated to obtain the first camera extrinsic parameters corresponding to each first camera.

[0020] Optionally, determining a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and a second pixel corresponding to the second camera image according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera includes:

[0021] Mapping each second pixel point in the second camera image to a second camera coordinate system according to the second camera intrinsic parameter to obtain second camera coordinate information corresponding to each second pixel point;

[0022] Mapping the second camera coordinate information corresponding to each second pixel point to a preset reference object coordinate system according to the second camera extrinsic parameter to obtain the reference object coordinate information corresponding to each second pixel point;

[0023] According to the first camera extrinsic parameters corresponding to each first camera, the reference object coordinate information corresponding to each second pixel point is mapped to the first camera coordinate system corresponding to each first camera, so as to obtain the first camera coordinate information corresponding to each second pixel point;

[0024] According to the distortion intrinsic parameter corresponding to each first camera, the first camera coordinate information of each second pixel point is converted to the first pixel point in the corresponding first camera image to obtain the target coordinate information of each second pixel point;

[0025] A pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is calculated according to the second camera coordinate information and the target coordinate information of each second pixel.

[0026] Optionally, generating a target image according to the target mapping matrix set includes:

[0027] According to each pixel mapping matrix in the target mapping matrix set, each first pixel in each first camera image is mapped to the second camera image to obtain an initial image corresponding to each first camera image;

[0028] All initial images are concatenated to obtain the target image.

[0029] Optionally, the acquiring the second camera parameter and determining the three-dimensional position information of each target object according to the second camera parameter and the two-dimensional detection frame information of each target object includes:

[0030] Obtaining target object type information and detection frame coordinate information in the two-dimensional detection frame information of each target object;

[0031] Determine the three-dimensional height information of each target object according to the target object type information;

[0032] Acquire a second camera parameter, and calculate the three-dimensional coordinate information of each target object according to the detection frame coordinate information and the second camera parameter;

[0033] The three-dimensional height information and three-dimensional coordinate information of each target object are combined to obtain the three-dimensional position information of each target object.

[0034] Optionally, determining the three-dimensional height information of each target object according to the target object type information includes:

[0035] According to the target type information, determining whether each target is a crossable type target;

[0036] If the target object is a crossable type target object, the three-dimensional height information of the corresponding target object is set to the first height value;

[0037] If the target object is a non-crossable type target object, the three-dimensional height information of the corresponding target object is set to a second height value, wherein the second height value is greater than the first height value.

[0038] Optionally, the acquiring the second camera parameters and calculating the three-dimensional coordinate information of each target object according to the detection frame coordinate information and the second camera parameters includes:

[0039] Acquire a second camera parameter, where the second camera parameter includes installation position information of the second camera;

[0040] According to the installation position information of the second camera, the detection frame coordinate information in the two-dimensional detection frame information of each target object is converted into three-dimensional coordinate information to obtain the three-dimensional coordinate information of each target object.

[0041] A second aspect of the present invention provides a device for determining a position of a target object, comprising:

[0042] An acquisition module, used for acquiring a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera;

[0043] A generating module, used for generating a target image according to the target mapping matrix set;

[0044] A detection module, used to perform target object detection on the target image to obtain two-dimensional detection frame information of at least one target object;

[0045] The determination module is used to obtain the second camera parameters and determine the three-dimensional position information of each target object according to the second camera parameters and the two-dimensional detection frame information of each target object.

[0046] Optionally, the acquisition module includes:

[0047] A first definition unit is used to define a second camera internal parameter and a second camera external parameter, where the second camera external parameter is used to indicate a relative position parameter of the second camera to a preset reference object;

[0048] A first acquisition unit is used to acquire a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, wherein the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to a second camera;

[0049] The first determining unit is used to determine a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and a second pixel corresponding to the second camera image according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, and generate a mapping matrix between each first camera image and the second camera image according to all pixel mapping matrices to obtain a target mapping matrix set.

[0050] Optionally, the first definition unit is specifically used to:

[0051] Setting a second camera intrinsic parameter, where the second camera intrinsic parameter is used to indicate an image parameter of the second camera;

[0052] A second camera coordinate system and a preset reference object coordinate system are defined, and relative position parameters from the second camera to the preset reference object are calculated according to the second camera coordinate system and the preset reference object coordinate system to obtain second camera extrinsics.

[0053] Optionally, the first acquiring unit is specifically used for:

[0054] Reading a distortion intrinsic parameter corresponding to each first camera of at least two first cameras, and defining a first camera coordinate system corresponding to each first camera;

[0055] According to the first camera coordinate system and the second camera coordinate system corresponding to each first camera, the relative position parameters from each first camera to the second camera are calculated to obtain the first camera extrinsic parameters corresponding to each first camera.

[0056] Optionally, the first determining unit is specifically configured to:

[0057] Mapping each second pixel point in the second camera image to a second camera coordinate system according to the second camera intrinsic parameter to obtain second camera coordinate information corresponding to each second pixel point;

[0058] Mapping the second camera coordinate information corresponding to each second pixel point to a preset reference object coordinate system according to the second camera extrinsic parameter to obtain the reference object coordinate information corresponding to each second pixel point;

[0059] According to the first camera extrinsic parameters corresponding to each first camera, the reference object coordinate information corresponding to each second pixel point is mapped to the first camera coordinate system corresponding to each first camera, so as to obtain the first camera coordinate information corresponding to each second pixel point;

[0060] According to the distortion intrinsic parameter corresponding to each first camera, the first camera coordinate information of each second pixel point is converted to the first pixel point in the corresponding first camera image to obtain the target coordinate information of each second pixel point;

[0061] A pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is calculated according to the second camera coordinate information and the target coordinate information of each second pixel.

[0062] Optionally, the generating module is specifically used for:

[0063] According to each pixel mapping matrix in the target mapping matrix set, each first pixel in each first camera image is mapped to the second camera image to obtain an initial image corresponding to each first camera image;

[0064] All initial images are concatenated to obtain the target image.

[0065] Optionally, the determining module includes:

[0066] A second acquisition unit is used to acquire target object type information and detection frame coordinate information in the two-dimensional detection frame information of each target object;

[0067] A second determining unit, configured to determine three-dimensional height information of each target object according to the target object type information;

[0068] A calculation unit, used to obtain a second camera parameter, and calculate the three-dimensional coordinate information of each target object according to the detection frame coordinate information and the second camera parameter;

[0069] The combining unit is used to combine the three-dimensional height information and the three-dimensional coordinate information of each target object to obtain the three-dimensional position information of each target object.

[0070] Optionally, the second determining unit is specifically configured to:

[0071] According to the target type information, determining whether each target is a crossable type target;

[0072] If the target object is a crossable type target object, the three-dimensional height information of the corresponding target object is set to the first height value;

[0073] If the target object is a non-crossable type target object, the three-dimensional height information of the corresponding target object is set to a second height value, wherein the second height value is greater than the first height value.

[0074] Optionally, the computing unit is specifically used for:

[0075] Acquire a second camera parameter, where the second camera parameter includes installation position information of the second camera;

[0076] According to the installation position information of the second camera, the detection frame coordinate information in the two-dimensional detection frame information of each target object is converted into three-dimensional coordinate information to obtain the three-dimensional coordinate information of each target object.

[0077] The third aspect of the present invention provides a device for determining the position of a target object, comprising: a memory and at least one processor, wherein the memory stores a computer program; the at least one processor calls the computer program in the memory so that the device for determining the position of a target object executes the above-mentioned method for determining the position of a target object.

[0078] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, which, when executed on a computer, enables the computer to execute the above-mentioned method for determining the position of a target object.

[0079] In the technical solution provided by the present invention, a target mapping matrix set is obtained, the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, one first camera image corresponds to one first camera; a target image is generated according to the target mapping matrix set; target object detection is performed on the target image to obtain two-dimensional detection frame information of at least one target object; second camera parameters are obtained, and three-dimensional position information of each target object is determined according to the second camera parameters and the two-dimensional detection frame information of each target object. In an embodiment of the present invention, each first camera image is converted to a second camera image through a mapping matrix between at least two first camera images and second camera images in the target mapping matrix set, a panoramic target image of the target perspective is obtained, and then target object detection is performed on the target image to obtain two-dimensional detection frame information of at least one target object, and the three-dimensional position information of each target object is calculated according to the second camera parameters. The present invention can improve the accuracy of target object position detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 A schematic diagram of an embodiment of a method for determining a position of a target object in an embodiment of the present invention;

[0081] Figure 2 is a schematic diagram of another embodiment of a method for determining a position of a target object in an embodiment of the present invention;

[0082] Figure 3 Schematic diagram of an embodiment of a device for determining a position of a target object in an embodiment of the present invention;

[0083] Figure 4 is a schematic diagram of another embodiment of a device for determining a position of a target object in an embodiment of the present invention;

[0084] Figure 5Schematic diagram of an embodiment of a device for determining the position of a target object in an embodiment of the present invention. DETAILED DESCRIPTION

[0085] Embodiments of the present invention provide a method, device, equipment and storage medium for determining a position of a target object, which are used to improve the accuracy of detecting the position of the target object.

[0086] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0087] It is understandable that the execution subject of the present invention may be a device for determining the position of a target object, or may be a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0088] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of a method for determining a target position in an embodiment of the present invention includes:

[0089] 10. Obtain a target mapping matrix set, where the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera;

[0090] It should be noted that there are at least two first cameras for collecting image information from different perspectives in a continuous spatial environment, such as monitoring cameras with different perspectives set at the entrance of a station, image acquisition cameras with different perspectives set on a self-moving robot, image acquisition cameras with different perspectives set on an autonomous driving vehicle, and violation monitoring cameras with different perspectives set on the same road space, etc., which are not specifically limited here. Each first camera corresponds to a first camera image, and the first camera images corresponding to each first camera are images collected by the camera at the same time.

[0091] It can be understood that the second camera image is used to indicate a virtual camera image of a continuous spatial environment at a target perspective, such as a left-view camera image of a station entrance, a rear-view camera image of a self-driving robot, a top-view camera image of an autonomous driving vehicle, a right-view camera image of the same road space, etc. The target mapping matrix set includes mapping matrices between at least two first camera images and the second camera images respectively. In one embodiment, the mapping matrix between each first camera image and the second camera image in the target mapping matrix set is used to indicate a perspective conversion relationship between the first camera image and the second camera image.

[0092] In one embodiment, each first camera can be a normal viewing angle camera or a wide-angle camera, such as a fisheye camera, a wide-angle lens, etc. Preferably, in order to collect image information in a larger range, each first camera is a fisheye camera, and the first camera image corresponding to each first camera is a fisheye camera image. Based on the above, since the image collected by the wide-angle camera has a certain degree of distortion, in this embodiment, the mapping matrix between each first camera image and the second camera image in the target mapping matrix set is also used to indicate the de-distortion viewing angle conversion relationship between the first camera image and the second camera image.

[0093] 20. Generate a target image according to the target mapping matrix set;

[0094] In one embodiment, the target image is used to indicate a panoramic image of a continuous spatial environment at a target perspective. Specifically, each first camera image is mapped to a second camera image according to a target mapping matrix set to obtain an initial image corresponding to each first camera image, and then all the initial images are stitched together to obtain the target image. Further, all the initial images are stitched together according to the same pixel blocks in the initial images to obtain the target image. For example, there are 4 first camera images, and the initial images corresponding to the first camera images are initial image A, initial image B, initial image C, and initial image D, respectively. Then, the same pixel blocks in the initial images include: the same pixel blocks ab in initial image A and initial image B, the same pixel blocks ac in initial image A and initial image C, the same pixel blocks ad in initial image A and initial image D, the same pixel blocks bc in initial image B and initial image C, the same pixel blocks bd in initial image B and initial image D, and the same pixel blocks cd in initial image C and initial image D.

[0095] In one embodiment, the target mapping matrix set includes a pixel mapping matrix between each first pixel in at least two first camera images and a second pixel in a second camera image. Then, step 20 includes: mapping each first pixel in each first camera image to the second camera image according to each pixel mapping matrix in the target mapping matrix set to obtain an initial image corresponding to each first camera image; and splicing all initial images to obtain a target image. Among them, each pixel mapping matrix in the target mapping matrix set is used to indicate the mapping relationship between the coordinate information of the corresponding pixel in the first camera image and the coordinate information of the pixel in the second camera image. The coordinate information can be two-dimensional coordinate information or three-dimensional coordinate information, which is not specifically limited here.

[0096] 30. Perform object detection on the target image to obtain two-dimensional detection frame information of at least one target object;

[0097] It can be understood that since the target image can observe the panoramic information of the continuous spatial environment from the target perspective, by performing target object detection on the target image, the problem of incomplete target object detection caused by the blind spot of vision when performing target detection through a single-perspective image can be avoided, thereby improving the accuracy of target object detection in the continuous spatial environment.

[0098] In one embodiment, step 30 includes: performing target detection on the target image through the trained target detection model to obtain two-dimensional detection frame information of at least one target, the target detection model includes a feature extraction network, a feature enhancement fusion network, a candidate frame extraction network and a candidate frame screening network, wherein the feature extraction network, the feature enhancement fusion network, the candidate frame extraction network and the candidate frame screening network are convolutional neural networks. Specifically, feature extraction, feature enhancement fusion, candidate frame extraction and candidate frame screening are performed on the target image through the trained target detection model to obtain two-dimensional detection frame information of at least one target. This embodiment can accurately identify targets in an image based on a deep learning model, and provides a data basis for subsequent target position calculation.

[0099] As an example but not limitation, the two-dimensional detection frame information of at least one target object is the minimum bounding rectangle of the target object, including but not limited to the detection frame coordinate information, width, height, rotation angle, existence information, target type information, etc. of the target object in the target image. For example, in an autonomous driving scenario, obstacle detection is performed on the target image to obtain the two-dimensional detection frame information of at least one obstacle.

[0100] 40. Obtain a second camera parameter, and determine the three-dimensional position information of each target object according to the second camera parameter and the two-dimensional detection frame information of each target object.

[0101] It can be understood that the three-dimensional position information is used to indicate the three-dimensional coordinate information of the target object, which is the relative position information relative to the preset reference object. As an example but not limitation, in an autonomous driving scenario, assuming that the vehicle body is the preset reference object, then the three-dimensional position information is the three-dimensional coordinate information relative to the vehicle body, that is, the position information of the target object in the three-dimensional coordinate system established by the preset reference object coordinate system with the vehicle body as the origin.

[0102] Based on the above, the three-dimensional position information of each target is calculated according to the second camera parameters and the two-dimensional detection frame information of each target, wherein the three-dimensional position information of each target may include the three-dimensional position information of all vertices of the target, and the three-dimensional position information of each target may also include the three-dimensional position information of the vertex of the target closest to a preset reference object, which is not specifically limited here.

[0103] In one embodiment, the three-dimensional position information of each target includes the three-dimensional coordinate information and the three-dimensional height information of the target. Specifically, step 40 includes: obtaining the target type information and the detection frame coordinate information in the two-dimensional detection frame information of each target; determining the three-dimensional height information of each target according to the target type information; obtaining the second camera parameter, and calculating the three-dimensional coordinate information of each target according to the detection frame coordinate information and the second camera parameter; combining the three-dimensional height information and the three-dimensional coordinate information of each target to obtain the three-dimensional position information of each target. In this embodiment, in order to improve the efficiency of determining the three-dimensional position of the target, the three-dimensional height information of each target can be directly determined according to the preset relationship between the target type information and the three-dimensional height information, for example, the height of the target of the non-obstacle type is set to 0, and the height of the target of the obstacle type is set to any value greater than 0, which is not specifically limited here. Then, the three-dimensional coordinate information of each target is calculated according to the second camera parameter and the detection frame coordinate information, wherein the second camera parameter can be used to indicate the relative position information of the second camera relative to the first camera, and can also be used to indicate the relative position information of the second camera relative to a preset reference object, such as the second camera parameter is used to indicate the relative position information of the second camera relative to the vehicle body.

[0104] Based on the above, as an example but not limited, in order to quickly determine the three-dimensional height of the target object, the three-dimensional height information of each target object is determined according to the target object type information, including: judging whether each target object is a crossable type target object according to the target object type information; if the target object is a crossable type target object, the three-dimensional height information of the corresponding target object is set to a first height value; if the target object is a non-crossable type target object, the three-dimensional height information of the corresponding target object is set to a second height value, wherein the second height value is greater than the first height value. For example, during the automatic driving process or the self-moving process of the self-moving robot, if the target object is identified as a plastic bag, and the target object type information of the plastic bag is a crossable type target object, then the three-dimensional height information of the target object is set to a first height value, such as 5 cm; if the target object is identified as a person, and the target object type information of the person is a non-crossable type target object, then the three-dimensional height information of the target object is set to a second height value greater than the first height value, such as 1 meter, and the specific value is not limited. This embodiment can obtain more detailed target object perception information so that the operating subject can make more accurate decisions.

[0105] Based on the above, the calculation process of the three-dimensional coordinate information of each target object is further described, the second camera parameters are obtained, and the three-dimensional coordinate information of each target object is calculated according to the detection frame coordinate information and the second camera parameters, including: obtaining the second camera parameters, the second camera parameters include the installation position information of the second camera; according to the installation position information of the second camera, the detection frame coordinate information in the two-dimensional detection frame information of each target object is converted into three-dimensional coordinate information to obtain the three-dimensional coordinate information of each target object. In this embodiment, the second camera parameters include the installation position information of the second camera, which is used to indicate the installation position information of the second camera relative to the preset reference object, wherein the preset reference object can be the first camera or other preset reference objects, such as setting the vehicle body as a preset parameter object in the autonomous driving scene. As an example and not a limitation, the installation position information includes but is not limited to installation height information and installation coordinate information. This embodiment can quickly calculate the three-dimensional coordinate information of the target object, thereby improving the efficiency of determining the position of the target object.

[0106] In an embodiment of the present invention, a mapping matrix between at least two first camera images and a second camera image is collected through a target mapping matrix, each first camera image is converted into a second camera image, a panoramic target image of a target viewing angle is obtained, and then target object detection is performed on the target image to obtain two-dimensional detection frame information of at least one target object, and three-dimensional position information of each target object is calculated according to the second camera parameters. The present invention can improve the accuracy of target object position detection.

[0107] See also Figure 2 , continue to explain the specific implementation of the above step 10:

[0108] 101. Define a second camera internal parameter and a second camera external parameter, where the second camera external parameter is used to indicate a relative position parameter of the second camera to a preset reference object;

[0109] It should be noted that in order to accurately obtain the location information of the target object, a second camera with a target perspective is defined, and the internal and external parameters of the second camera are set, so as to subsequently map the first camera image to the second camera to obtain a panoramic image of the target perspective, and perform target detection on the panoramic image to obtain complete target object information. It can be understood that the second camera can be a real camera installed at a preset perspective, or it can be a virtual camera with a preset perspective that is not actually installed, and no specific limitation is made here.

[0110] As an example and not a limitation, the second camera intrinsic parameter is used to indicate the parameters of the second camera image, such as the width, height, focal length and other values ​​of the image. In one embodiment, the second camera intrinsic parameter is an intrinsic parameter matrix. The second camera extrinsic parameter is used to indicate the relative position parameter of the second camera to a preset reference object, wherein the preset reference object can be any one of the first cameras or other reference objects. For example, the second camera extrinsic parameter can be used to indicate the relative position parameter of the second camera to the vehicle body. The relative position parameter is a real position parameter in three-dimensional space. For example, the relative position parameter includes a horizontal axis offset value and a vertical axis offset value from the second camera to the vehicle body.

[0111] Specifically, step 101 includes: setting the second camera internal parameter, which is used to indicate the image parameters of the second camera; defining the second camera coordinate system and the preset reference object coordinate system, and calculating the relative position parameters from the second camera to the preset reference object according to the second camera coordinate system and the preset reference object coordinate system, to obtain the second camera external parameter. In this embodiment, by defining the second camera coordinate system and the preset reference object coordinate system, calculating the relative position parameters from the second camera to the preset reference object, the reference standard can be obtained through a unified origin coordinate, so that the accuracy of determining the position of the target object is improved.

[0112] 102. Obtain a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, where the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to the second camera;

[0113] As an example but not a limitation, when the first camera is a wide-angle camera, the captured image has a certain degree of distortion. Therefore, the distortion intrinsic parameters corresponding to each of the at least two first cameras are obtained, and the distortion intrinsic parameters are used to determine the subsequent mapping matrix, and the first camera extrinsic parameters corresponding to each of the at least two first cameras are used to indicate the relative position parameters of each first camera to the second camera, wherein the relative position parameters are real position parameters in three-dimensional space, such as the installation position distance between the first camera and the second camera is 1 meter, and the relative position parameters of each first camera to the second camera include the horizontal axis offset value and the vertical axis offset value of each first camera to the second camera. In one embodiment, the first camera extrinsic parameters corresponding to each first camera are matrices.

[0114] Specifically, step 102 includes: reading the distortion intrinsic parameters corresponding to each of the at least two first cameras, and defining the first camera coordinate system corresponding to each first camera; calculating the relative position parameters from each first camera to the second camera according to the first camera coordinate system and the second camera coordinate system corresponding to each first camera, and obtaining the first camera extrinsic parameters corresponding to each first camera. In this embodiment, the distortion intrinsic parameters corresponding to each first camera are preset parameters of the camera, which can be obtained by directly reading and will not change due to changes in the external environment. Then, by defining the first camera coordinate system corresponding to each first camera, the relative position parameters corresponding to each first camera are calculated, and the first camera extrinsic parameters corresponding to each first camera are obtained, which can improve the accuracy of the target position determination through data standardization.

[0115] 103. Determine a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and a second pixel in the second camera image corresponding to each first camera according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, and generate a mapping matrix between each first camera image and the second camera image according to all pixel mapping matrices to obtain a target mapping matrix set;

[0116] It should be noted that since the second camera intrinsic parameters, the second camera extrinsic parameters, and the distortion intrinsic parameters and the first camera extrinsic parameters corresponding to each first camera will no longer change after being determined, they are all constants. Therefore, the pixel mapping matrix and the target mapping matrix set calculated by these constants are also constants. As long as the installation positions of the first camera and the second camera do not change or the camera types are not changed, the target mapping matrix set obtained by one calculation can be stored as a constant. Each time the position of the target object is determined, it is only necessary to directly read the constant without repeated calculation, which greatly improves the efficiency of target object position determination.

[0117] In one implementation, since the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera are known, the second camera image can be mapped to the first camera image, and the first camera image can be mapped to the second camera image to establish a mapping relationship between each pixel in the second camera image and the pixel of the first camera image, that is, to obtain a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image. In one implementation, the specific implementation of mapping the first camera image to the second camera image is as follows: according to the distortion intrinsic parameter and the first camera intrinsic parameter corresponding to each first camera, each pixel in the first camera image corresponding to each first camera is mapped to the first camera image; The first pixel point is mapped to the first camera coordinate system to obtain the first camera coordinate information corresponding to each first pixel point; according to the first camera extrinsic parameter corresponding to each first camera, the first camera coordinate information corresponding to each first pixel point is mapped to the second camera coordinate system to obtain the second camera coordinate information corresponding to each first pixel point; according to the second camera extrinsic parameter and the second camera intrinsic parameter, the second camera coordinate system corresponding to each first pixel point is converted to the corresponding second pixel point in the second camera image to obtain the target coordinate information corresponding to each first pixel point; according to the first camera coordinate information and the target coordinate information corresponding to each first pixel point, a pixel mapping matrix between each first pixel point in the first camera image corresponding to each first camera and the second pixel point corresponding to the second camera image is calculated.

[0118] Furthermore, a specific implementation method of mapping the second camera image to the first camera image is further described, that is, according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is determined, including: according to the second camera intrinsic parameter, mapping each second pixel in the second camera image to the second camera coordinate system to obtain the second camera coordinate information corresponding to each second pixel; according to the second camera extrinsic parameter, mapping the second camera coordinate information corresponding to each second pixel to a preset reference object coordinate system to obtain each second pixel Corresponding reference object coordinate information; according to the first camera external parameters corresponding to each first camera, mapping the reference object coordinate information corresponding to each second pixel point to the first camera coordinate system corresponding to each first camera, obtaining the first camera coordinate information corresponding to each second pixel point; according to the distortion internal parameters corresponding to each first camera, converting the first camera coordinate information of each second pixel point to the first pixel point in the corresponding first camera image, obtaining the target coordinate information of each second pixel point; according to the second camera coordinate information and the target coordinate information of each second pixel point, calculating the pixel mapping matrix between each first pixel point in the first camera image corresponding to each first camera and the second pixel point corresponding to the second camera image. This embodiment can omit the generation process of the distorted image, directly calculate the mapping relationship between the distorted image and the panoramic image of the target perspective, so that the efficiency of generating the panoramic image of the target perspective is improved, thereby improving the efficiency and accuracy of determining the position of the target object through the panoramic image.

[0119] In an embodiment of the present invention, a mapping matrix between at least two first camera images and the second camera images in a target mapping matrix set is determined through a second camera intrinsic parameter, a second camera extrinsic parameter, and a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each first camera, each first camera image is converted into a second camera image, a panoramic target image of a target viewing angle is obtained, and then target object detection is performed on the target image to obtain two-dimensional detection frame information of at least one target object, and three-dimensional position information of each target object is calculated according to the second camera parameters. The present invention can improve the accuracy of target object position detection.

[0120] The above describes the method for determining the position of a target object in the embodiment of the present invention. The following describes the device for determining the position of a target object in the embodiment of the present invention. Figure 3 In one embodiment of the present invention, a device for determining a target position includes:

[0121] An acquisition module 301 is used to acquire a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera;

[0122] A generating module 302, configured to generate a target image according to the target mapping matrix set;

[0123] A detection module 303 is used to perform target object detection on the target image to obtain two-dimensional detection frame information of at least one target object;

[0124] The determination module 304 is used to obtain second camera parameters, and determine the three-dimensional position information of each target object according to the second camera parameters and the two-dimensional detection frame information of each target object.

[0125] In an embodiment of the present invention, a mapping matrix between at least two first camera images and a second camera image is collected through a target mapping matrix, each first camera image is converted into a second camera image, a panoramic target image of a target viewing angle is obtained, and then target object detection is performed on the target image to obtain two-dimensional detection frame information of at least one target object, and three-dimensional position information of each target object is calculated according to the second camera parameters. The present invention can improve the accuracy of target object position detection.

[0126] See also Figure 4 Another embodiment of the device for determining the position of a target object in the embodiment of the present invention includes:

[0127] An acquisition module 301 is used to acquire a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera;

[0128] A generating module 302, configured to generate a target image according to the target mapping matrix set;

[0129] A detection module 303 is used to perform target object detection on the target image to obtain two-dimensional detection frame information of at least one target object;

[0130] The determination module 304 is used to obtain second camera parameters, and determine the three-dimensional position information of each target object according to the second camera parameters and the two-dimensional detection frame information of each target object.

[0131] Optionally, the acquisition module 301 includes:

[0132] A first definition unit 3011 is used to define a second camera intrinsic parameter and a second camera extrinsic parameter, where the second camera extrinsic parameter is used to indicate a relative position parameter of the second camera to a preset reference object;

[0133] A first acquisition unit 3012 is used to acquire a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, wherein the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to the second camera;

[0134] The first determination unit 3013 is used to determine a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and a second pixel corresponding to the second camera image according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, and generate a mapping matrix between each first camera image and the second camera image according to all pixel mapping matrices to obtain a target mapping matrix set.

[0135] Optionally, the first definition unit 3011 is specifically used to:

[0136] Setting a second camera intrinsic parameter, where the second camera intrinsic parameter is used to indicate an image parameter of the second camera;

[0137] A second camera coordinate system and a preset reference object coordinate system are defined, and relative position parameters from the second camera to the preset reference object are calculated according to the second camera coordinate system and the preset reference object coordinate system to obtain second camera extrinsics.

[0138] Optionally, the first acquiring unit 3012 is specifically configured to:

[0139] Reading a distortion intrinsic parameter corresponding to each first camera of at least two first cameras, and defining a first camera coordinate system corresponding to each first camera;

[0140] According to the first camera coordinate system and the second camera coordinate system corresponding to each first camera, the relative position parameters from each first camera to the second camera are calculated to obtain the first camera extrinsic parameters corresponding to each first camera.

[0141] Optionally, the first determining unit 3013 is specifically configured to:

[0142] Mapping each second pixel point in the second camera image to a second camera coordinate system according to the second camera intrinsic parameter to obtain second camera coordinate information corresponding to each second pixel point;

[0143] Mapping the second camera coordinate information corresponding to each second pixel point to a preset reference object coordinate system according to the second camera extrinsic parameter to obtain the reference object coordinate information corresponding to each second pixel point;

[0144] According to the first camera extrinsic parameters corresponding to each first camera, the reference object coordinate information corresponding to each second pixel point is mapped to the first camera coordinate system corresponding to each first camera, so as to obtain the first camera coordinate information corresponding to each second pixel point;

[0145] According to the distortion intrinsic parameter corresponding to each first camera, the first camera coordinate information of each second pixel point is converted to the first pixel point in the corresponding first camera image to obtain the target coordinate information of each second pixel point;

[0146] A pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is calculated according to the second camera coordinate information and the target coordinate information of each second pixel.

[0147] Optionally, the generating module 302 is specifically used for:

[0148] According to each pixel mapping matrix in the target mapping matrix set, each first pixel in each first camera image is mapped to the second camera image to obtain an initial image corresponding to each first camera image;

[0149] All initial images are concatenated to obtain the target image.

[0150] Optionally, the determining module 304 includes:

[0151] The second acquisition unit 3041 is used to acquire the target object type information and the detection frame coordinate information in the two-dimensional detection frame information of each target object;

[0152] A second determining unit 3042 is used to determine the three-dimensional height information of each target object according to the target object type information;

[0153] A calculation unit 3043 is used to obtain a second camera parameter, and calculate the three-dimensional coordinate information of each target object according to the detection frame coordinate information and the second camera parameter;

[0154] The combining unit 3044 is used to combine the three-dimensional height information and the three-dimensional coordinate information of each target object to obtain the three-dimensional position information of each target object.

[0155] Optionally, the second determining unit 3042 is specifically configured to:

[0156] According to the target type information, determining whether each target is a crossable type target;

[0157] If the target object is a crossable type target object, the three-dimensional height information of the corresponding target object is set to the first height value;

[0158] If the target object is a non-crossable type target object, the three-dimensional height information of the corresponding target object is set to a second height value, wherein the second height value is greater than the first height value.

[0159] Optionally, the calculation unit 3043 is specifically used for:

[0160] Acquire a second camera parameter, where the second camera parameter includes installation position information of the second camera;

[0161] According to the installation position information of the second camera, the detection frame coordinate information in the two-dimensional detection frame information of each target object is converted into three-dimensional coordinate information to obtain the three-dimensional coordinate information of each target object.

[0162] By using the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, the mapping matrices between at least two first camera images in the target mapping matrix set and the second camera images are determined, each first camera image is converted to a second camera image, a panoramic target image of the target viewing angle is obtained, and then target object detection is performed on the target image to obtain two-dimensional detection frame information of at least one target object, and three-dimensional position information of each target object is calculated according to the second camera parameters. The present invention can improve the accuracy of target object position detection.

[0163] above Figure 3 and Figure 4 The device for determining the position of a target object in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The device for determining the position of a target object in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0164] Figure 5 1 is a schematic diagram of the structure of a device for determining the position of a target object provided by an embodiment of the present invention. The device 500 for determining the position of a target object may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 may be temporary storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of computer program operations in the device 500 for determining the position of the target object. Furthermore, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of computer program operations in the storage medium 530 on the device 500 for determining the position of the target object.

[0165] The target position determination device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 5 The structure of the target position determination device shown does not constitute a limitation on the target position determination device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0166] The present invention also provides a computer device, which includes a memory and a processor, wherein a computer-readable computer program is stored in the memory, and when the computer-readable computer program is executed by the processor, the processor executes the steps of the method for determining the position of the target object in the above-mentioned embodiments.

[0167] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. A computer program is stored in the computer-readable storage medium. When the computer program runs on a computer, the computer executes the steps of the method for determining the position of the target object.

[0168] Furthermore, the computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the blockchain node, etc.

[0169] The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0170] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0171] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several computer programs to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0172] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for determining the position of a target object, characterized in that: The method for determining the position of the target object includes: Acquire a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera; generating a target image according to the target mapping matrix set; Performing target object detection on the target image to obtain two-dimensional detection frame information of at least one target object; Obtaining target object type information and detection frame coordinate information in the two-dimensional detection frame information of each target object; Determine the three-dimensional height information of each target object according to the target object type information; Acquire a second camera parameter, and calculate the three-dimensional coordinate information of each target object according to the detection frame coordinate information and the second camera parameter; The three-dimensional height information and three-dimensional coordinate information of each target object are combined to obtain the three-dimensional position information of each target object.

2. The method for determining the position of a target object according to claim 1, characterized in that: The target mapping matrix set is obtained, and the target mapping matrix set includes mapping matrices between at least two first camera images and the second camera image, respectively, including: defining a second camera intrinsic parameter and a second camera extrinsic parameter, wherein the second camera extrinsic parameter is used to indicate a relative position parameter of the second camera to a preset reference object; Obtaining a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, wherein the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to the second camera; A pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is determined according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera, and a mapping matrix between each first camera image and the second camera image is generated according to all pixel mapping matrices to obtain a target mapping matrix set.

3. The method for determining the position of a target object according to claim 2, characterized in that: The defining of the second camera intrinsic parameter and the second camera extrinsic parameter, wherein the second camera extrinsic parameter is used to indicate a relative position parameter of the second camera to a preset reference object, includes: Setting a second camera intrinsic parameter, where the second camera intrinsic parameter is used to indicate an image parameter of the second camera; A second camera coordinate system and a preset reference object coordinate system are defined, and relative position parameters from the second camera to the preset reference object are calculated according to the second camera coordinate system and the preset reference object coordinate system to obtain second camera extrinsics.

4. The method for determining the position of a target object according to claim 3, characterized in that: The obtaining of a distortion intrinsic parameter and a first camera extrinsic parameter corresponding to each of the at least two first cameras, wherein the first camera extrinsic parameter corresponding to each first camera is used to indicate a relative position parameter from each first camera to the second camera, includes: Reading a distortion intrinsic parameter corresponding to each first camera of at least two first cameras, and defining a first camera coordinate system corresponding to each first camera; According to the first camera coordinate system and the second camera coordinate system corresponding to each first camera, the relative position parameters from each first camera to the second camera are calculated to obtain the first camera extrinsic parameters corresponding to each first camera.

5. The method for determining the position of a target object according to claim 2, characterized in that: The step of determining a pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and a second pixel corresponding to the second camera image according to the second camera intrinsic parameter, the second camera extrinsic parameter, and the distortion intrinsic parameter and the first camera extrinsic parameter corresponding to each first camera includes: Mapping each second pixel point in the second camera image to a second camera coordinate system according to the second camera intrinsic parameter to obtain second camera coordinate information corresponding to each second pixel point; Mapping the second camera coordinate information corresponding to each second pixel point to a preset reference object coordinate system according to the second camera extrinsic parameter to obtain the reference object coordinate information corresponding to each second pixel point; According to the first camera extrinsic parameters corresponding to each first camera, the reference object coordinate information corresponding to each second pixel point is mapped to the first camera coordinate system corresponding to each first camera, so as to obtain the first camera coordinate information corresponding to each second pixel point; According to the distortion intrinsic parameter corresponding to each first camera, the first camera coordinate information of each second pixel point is converted to the first pixel point in the corresponding first camera image to obtain the target coordinate information of each second pixel point; A pixel mapping matrix between each first pixel in the first camera image corresponding to each first camera and the second pixel corresponding to the second camera image is calculated according to the second camera coordinate information and the target coordinate information of each second pixel.

6. The method for determining the position of a target object according to claim 2, characterized in that: Generating a target image according to the target mapping matrix set comprises: According to each pixel mapping matrix in the target mapping matrix set, each first pixel in each first camera image is mapped to the second camera image to obtain an initial image corresponding to each first camera image; All initial images are concatenated to obtain the target image.

7. The method for determining the position of a target object according to claim 1, characterized in that: The determining the three-dimensional height information of each target object according to the target object type information includes: According to the target type information, determining whether each target is a crossable type target; If the target object is a crossable type target object, the three-dimensional height information of the corresponding target object is set to the first height value; If the target object is a non-crossable type target object, the three-dimensional height information of the corresponding target object is set to a second height value, wherein the second height value is greater than the first height value.

8. The method for determining the position of a target object according to claim 1, characterized in that: The acquiring the second camera parameters and calculating the three-dimensional coordinate information of each target object according to the detection frame coordinate information and the second camera parameters includes: Acquire a second camera parameter, where the second camera parameter includes installation position information of the second camera; According to the installation position information of the second camera, the detection frame coordinate information in the two-dimensional detection frame information of each target object is converted into three-dimensional coordinate information to obtain the three-dimensional coordinate information of each target object.

9. A device for determining the position of a target object, characterized in that: The device for determining the position of the target object comprises: An acquisition module, used for acquiring a target mapping matrix set, wherein the target mapping matrix set includes mapping matrices between at least two first camera images and second camera images, respectively, and one first camera image corresponds to one first camera; A generating module, used for generating a target image according to the target mapping matrix set; A detection module, used to perform target object detection on the target image to obtain two-dimensional detection frame information of at least one target object; A determination module is used to obtain target type information and detection frame coordinate information in the two-dimensional detection frame information of each target; determine the three-dimensional height information of each target according to the target type information; obtain a second camera parameter, and calculate the three-dimensional coordinate information of each target according to the detection frame coordinate information and the second camera parameter; and obtain the three-dimensional position information of each target by combining the three-dimensional height information and the three-dimensional coordinate information of each target.

10. A device for determining the position of a target object, characterized in that: The target position determination device comprises: a memory and at least one processor, wherein the memory stores a computer program; The at least one processor calls the computer program in the memory to enable the target position determination device to perform the target position determination method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for determining the position of a target object as claimed in any one of claims 1 to 8 is implemented.