Grid optimization-based coordinate mapping thermodynamic diagram generation method and device, equipment and medium

Abstract

The invention discloses a coordinate mapping thermodynamic diagram generation method and device based on grid optimization, equipment and a medium, and the method comprises the steps: carrying out the grid drawing of a target image after the calibration of a camera, and mapping the grid to a world coordinate, and enabling the target image to be obtained through the shooting of the camera; detecting a foot point of the first target based on deep learning, and acquiring an image coordinate of the foot point in the target picture; determining a first grid to which the foot point belongs based on the image coordinate of the foot point; mapping the image coordinates of the foot points to obtain world coordinates of the foot points; and determining a real mapping point location of the foot point according to a relationship between the first grid and the world coordinate of the foot point. The invention belongs to the field of computer vision. According to the invention, the mapping precision from the image coordinates to the world coordinate system can be improved.

Description

Technical Field

[0001] This invention relates to the field of computer vision, and more particularly to a method, apparatus, device, and medium for generating coordinate mapping heatmaps based on grid optimization. Background Technology

[0002] With the rapid development of computer vision technology, object detection and spatial coordinate mapping have become indispensable core technologies in modern science and technology. They play a crucial role in numerous high-tech applications such as autonomous driving, intelligent monitoring, intelligent robots, augmented reality, and drone navigation. Especially in the field of autonomous driving, accurate object localization and spatial coordinate mapping are not only the foundation of vehicle environmental perception but also key to achieving safe driving, path planning, obstacle avoidance, and intelligent decision-making. Precise object localization and mapping help the system understand and predict changes in the environment, thereby making more intelligent and rational decisions.

[0003] Currently, accurately mapping target information (such as the pixel coordinates of an object in an image) obtained from a monocular camera to a real-world 3D coordinate system and generating a corresponding heatmap for visualization remains a technical challenge. Therefore, this invention proposes a coordinate mapping heatmap generation method based on mesh optimization. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for generating coordinate mapping heatmaps based on mesh optimization. This solves the technical problem in the prior art of accurately mapping target information obtained from a monocular camera to a three-dimensional coordinate system in the real world, thereby improving the accuracy of the mapping.

[0005] In a first aspect, the present invention provides a method for generating coordinate mapping heatmaps based on mesh optimization, the method comprising:

[0006] After the camera is calibrated, a grid is drawn for the target image and mapped to world coordinates. The target image is captured by the camera.

[0007] Based on deep learning, the foot points of the first target are detected, and the image coordinates of the foot points in the target image are obtained;

[0008] Based on the image coordinates of the foot point, determine the first grid to which the foot point belongs;

[0009] Map the image coordinates of the foot point to obtain the world coordinates of the foot point;

[0010] Based on the relationship between the first grid and the world coordinates of the foot point, determine the actual mapped position of the foot point.

[0011] Furthermore, based on the relationship between the world coordinates of the first grid and the foot point, the actual mapped position of the foot point is determined, including:

[0012] Determine if the world coordinates of the foot point are within the world coordinates corresponding to the first grid.

[0013] If the foot is in the world coordinates, then the true mapping point of the foot is the world coordinates of the foot; if the foot is not in the world coordinates, then in the world coordinates, a line is drawn connecting the center of the first grid and the foot, and the intersection of the line and the first grid is taken as the true mapping point of the foot.

[0014] Furthermore, it also includes:

[0015]

[0016] Where P is the actual mapped position of the foot point, G is the first grid, and p aut G represents the world coordinates of the foot point. edge The boundary of the first grid, p ma The coordinates of the foot point are shown in the image.

[0017] Furthermore, camera calibration includes:

[0018]

[0019] Where, k 3×3 R is the intrinsic parameter matrix of the camera. 3×3 Let t be a rotation matrix. 3×1 The vector is a translation vector, where u and v represent image coordinates, x, y, and h are world coordinates, and z is the translation vector. c This is for depth information.

[0020] Furthermore, several objectives are unified onto a single plane, including:

[0021]

[0022] Where, r 11 to r 33 All are elements in the rotation matrix, z is the depth parameter, t1 to t3 are translation vectors, and f x f y u0 and v0 are all elements in the intrinsic parameter matrix.

[0023] Furthermore, camera calibration also includes:

[0024] Configure the camera's intrinsic parameters, including: setting the camera's focal length, principal point, or distortion coefficient;

[0025] Configure the camera's extrinsic parameters, including setting the camera's rotation matrix and translation vector.

[0026] Furthermore, based on deep learning, the footpoints for detecting the first objective include:

[0027] The neural network to be trained is trained based on the YOLOv5 algorithm to obtain a human detection model. The human detection model is used to identify several targets in the target image and configure a followable bounding box for each target.

[0028] The positions of several key points of the target are obtained based on a preset attitude estimation model;

[0029] Extract the target's foot points from several key locations.

[0030] Secondly, the present invention provides a coordinate mapping heatmap generation device based on mesh optimization, the device comprising:

[0031] The mapping module is used to draw a grid on the target image after the camera is calibrated, and to map the grid to world coordinates. The target image is captured by the camera.

[0032] The acquisition module is used to detect the foot points of the first target based on deep learning and obtain the image coordinates of the foot points in the target image.

[0033] The first grid module is used to determine the first grid to which a foot belongs based on its image coordinates.

[0034] The world coordinates module is used to map the image coordinates of the foot point to obtain the world coordinates of the foot point;

[0035] The mapping point module is used to determine the actual mapping point of the foot point based on the relationship between the world coordinates of the first grid and the foot point.

[0036] Thirdly, the present invention provides an electronic device, comprising:

[0037] processor;

[0038] Memory used to store processor-executable instructions;

[0039] The processor is configured to execute a grid-optimized coordinate mapping heatmap generation method as provided in the first aspect.

[0040] Fourthly, the present invention provides a non-transitory computer-readable storage medium that, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the grid-optimized coordinate mapping heatmap generation method provided in the first aspect above.

[0041] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0042] This invention first uses automatic mapping technology to quickly obtain preliminary world coordinate mapping results, and then introduces a manual correction step on this basis, allowing users to fine-tune key points according to the actual situation. This invention makes full use of the speed advantage of automatic mapping, while improving the positioning accuracy of distant targets with the help of manual correction, ensuring that the generated heat map is smoother and more accurate.

[0043] This invention utilizes the intrinsic parameter matrix of the camera to perform distortion correction and back-projection processing on image coordinates, converting them into normalized camera coordinates. Then, based on known planar information (e.g., ground or tabletop), a fixed depth value (z-coordinate) is set to determine the actual 3D position of each target in the camera coordinate system. The extrinsic parameters of the camera (rotation matrix and translation vector) are then applied to convert these camera coordinates into world coordinates. This invention simplifies the complexity of depth estimation and can also efficiently generate heatmaps. Attached Figure Description

[0044] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A schematic flowchart illustrating the coordinate mapping heatmap generation method based on mesh optimization provided by this invention;

[0046] Figure 2 A schematic diagram of the grid provided by the present invention;

[0047] Figure 3 This is a schematic diagram illustrating the relationship between the mesh and the foot points provided by the present invention. Detailed Implementation

[0048] The embodiments of the present invention provide a coordinate mapping heatmap generation method based on mesh optimization, which solves the technical problem in the prior art that it is difficult to accurately map target information obtained from a monocular camera to the three-dimensional coordinate system of the real world.

[0049] The technical solution of this invention is to solve the above-mentioned technical problems, and the overall idea is as follows:

[0050] A grid-optimized coordinate mapping heatmap generation method includes: after camera calibration, drawing a grid on the target image and mapping the grid to world coordinates, wherein the target image is captured by the camera; detecting the foot points of the first target based on deep learning and obtaining the image coordinates of the foot points in the target image; determining the first grid to which the foot point belongs based on the image coordinates of the foot point; mapping the image coordinates of the foot point to obtain the world coordinates of the foot point; and determining the actual mapped position of the foot point according to the relationship between the first grid and the world coordinates of the foot point.

[0051] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0052] First, it should be clarified that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0053] The present invention is described as follows: Depth information estimation based on a fixed plane includes:

[0054]

[0055] Where, k 3×3 R is the intrinsic parameter matrix of the camera. 3×3 Let t be a rotation matrix. 3×1 The vector is a translation vector, where u and v represent image coordinates, x, y, and h are world coordinates, and z is the translation vector. c As depth information is, understandably, difficult to acquire using conventional cameras.

[0056] Therefore, several targets can be unified onto a single plane (in this invention, "target" refers to a person), and a partial heatmap of these targets can be generated to create a scene. This means that h can be used as a constant, and depth information z... c It can be represented by x, y, and several constants. Expanding the matrix, we get the following formula:

[0057]

[0058] Where, r 11 to r 33 All are elements in the rotation matrix, z is the depth parameter, t1 to t3 are translation vectors, and f x f y u0 and v0 are all elements in the intrinsic parameter matrix.

[0059] Expanding the above formula, we get:

[0060]

[0061] When h is a constant, it is a system of three linear equations in three variables, which can be solved for x, y, and z (the premise of the solution is that all targets are assumed to be on the same plane, which conforms to the preset of the heat map).

[0062] However, it should be emphasized that in order to ensure that the two-dimensional plane captured by the camera is on the same horizontal plane, it is necessary to detect the foot points of the person.

[0063] According to the inventor's tests, in the above mapping, the error is relatively small at close range (within 6 meters), between 0.2 and 0.4, which is within an acceptable range. However, the error increases rapidly at longer distances: at 10 meters, the error is about 2 meters; at 15 meters, the error reaches 4 meters.

[0064] Therefore, it can be concluded that automatic mapping is more accurate at close range, but at long range, the accuracy of automatic mapping is not as good as manual mapping (i.e., drawing a grid area and then determining the grid to which the target belongs). Therefore, in order to improve the mapping effect, the present invention provides steps S11-S15.

[0065] This invention provides, for example Figure 1 The grid-optimized coordinate mapping heatmap generation method shown includes steps S11-S15:

[0066] Step S11: After the camera is calibrated, a grid is drawn on the target image and mapped to world coordinates. The target image is captured by the camera, and the size of the grid is consistent in world coordinates.

[0067] Camera calibration is the process of determining the geometric parameters of a camera, mainly including intrinsic parameters (such as focal length, principal point coordinates, distortion coefficients, etc.) and extrinsic parameters (such as rotation matrix and translation vector). The main purpose of calibration is to establish a camera imaging model so that points in three-dimensional space can be accurately projected onto a two-dimensional image plane, or conversely, the position of a point in two-dimensional image can be deduced from its position in three-dimensional space.

[0068] like Figure 2 As shown, after completing camera calibration, a virtual mesh can be drawn on the captured target image. The mesh can help understand or verify the effect of camera calibration and can also be used for subsequent 3D reconstruction or other computer vision tasks.

[0069] Then, the grid on the image plane can be mapped to the world coordinate system (using the intrinsic and extrinsic parameters obtained from camera calibration to perform coordinate transformation).

[0070] The specific steps include: defining the position of grid points in the image coordinate system, transforming from the image coordinate system (also known as the image coordinate system) to the camera coordinate system, and using the camera's extrinsic parameters (rotation matrix and translation vector) to transform the points in the camera coordinate system to the world coordinate system.

[0071] Step S12: Detect the foot points of the first target based on deep learning, and obtain the image coordinates of the foot points in the target image.

[0072] The method for detecting the foot points of the first target based on deep learning includes: training the neural network to be trained based on the YOLOv5 algorithm to obtain a human detection model, wherein the human detection model is used to identify several targets in the target image and configure a followable bounding box for each target; obtaining several key point positions of the target based on a preset pose estimation model; and extracting the foot points of the target from the several key point positions.

[0073] YOLOv5 is an object detection algorithm, belonging to the family of single-stage object detection algorithms. Based on the Convolutional Neural Network (CNN) architecture in deep learning, YOLOv5 can quickly process and analyze input image or video data. The core idea of ​​the YOLOv5 algorithm is to transform the object detection task into a regression problem, directly predicting the bounding boxes and class probabilities of objects at multiple scales of the image. YOLOv5 employs techniques such as Feature Pyramid Network (FPN) to fuse feature information from different levels, improving the detection accuracy for objects of different sizes. During training, YOLOv5 uses a large-scale labeled dataset for learning and adjusts the network parameters through backpropagation, enabling the model to accurately identify and locate various objects. It achieves a good balance between speed and accuracy, enabling it to quickly and accurately detect target objects in scenarios with high real-time requirements, such as intelligent security, autonomous driving, and robot vision, providing crucial visual information support for subsequent decision-making and actions.

[0074] In addition to the methods provided above, the present invention also provides the following methods, including:

[0075] Collect and label a dataset for foot points, where the foot point positions of each person in each image need to be labeled; modify the YOLOv5 configuration file to adapt to the new task requirements, fine-tune the YOLOv5 model using the dataset; evaluate the model performance on the test set, and further optimize the model based on the results.

[0076] Also includes:

[0077] Design a network architecture that branches out into two subnetworks based on a shared feature extractor: one for object detection (such as YOLOv5) and the other for keypoint estimation.

[0078] Use a dataset containing object bounding boxes and keypoint annotations for training; ensure that the loss function can take into account the performance of both tasks simultaneously.

[0079] During inference, the model outputs both bounding boxes and keypoint locations, from which the required footpoint information can be directly extracted.

[0080] Step S13: Determine the first grid to which the foot belongs based on the image coordinates of the foot.

[0081] The image is divided into several non-overlapping rectangular regions according to a predetermined grid size. For example, if the image resolution is 1920×1080 pixels and a grid size of 50×50 pixels is selected, a grid of 38×21.6 pixels will be obtained.

[0082] A footpoint refers to the position of a specific object (the foot of a pedestrian) in an image. For each identified object, the two-dimensional coordinates of its footpoint on the target image can be obtained. Image coordinates are given in pixels (therefore, they can also be called pixel coordinates).

[0083] Once the image coordinates of the foot points are determined, the grid index can be calculated:

[0084] For the horizontal direction (X-axis), grid index , where width is the width of a single grid cell.

[0085] For the vertical direction (Y-axis), grid index Where height is the height of a single grid cell. All values ​​can be rounded down.

[0086] Step S14: Map the image coordinates of the foot point to obtain the world coordinates of the foot point.

[0087] Specifically, it includes:

[0088] The image coordinates are distorted and back-projected based on the camera's intrinsic parameter matrix and distortion coefficients, and then converted into normalized camera coordinates. Then, the actual 3D camera coordinates are determined based on the depth information in the scene (set as a constant). Finally, the camera's extrinsic parameters (rotation matrix and translation vector) are applied to convert the camera coordinates into world coordinates.

[0089] Step S15: Determine the actual mapped position of the foot point based on the relationship between the world coordinates of the first grid and the foot point.

[0090] Based on the relationship between the first grid and the world coordinates of the foot point, determine the actual mapped position of the foot point, including:

[0091] Determine if the world coordinates of the foot point are within the world coordinates corresponding to the first grid.

[0092] If it is in a certain position, then the world coordinates of the foot point are used as the actual mapped position of the foot point; for example... Figure 3 As shown (A is the grid center, B is the foot point, and C is the intersection point), if it is not in the grid, then in world coordinates, draw a line connecting the center of the first grid and the foot point, and take the intersection of the line and the first grid as the actual mapped point of the foot point.

[0093] Also includes:

[0094]

[0095] Where P is the actual mapped position of the foot point, G is the first grid, and p aut G represents the world coordinates of the foot point. edge The boundary of the first grid, p ma The coordinates of the foot point are shown in the image.

[0096] In summary, this invention provides a coordinate mapping heatmap generation method based on grid optimization. The method includes: after camera calibration, drawing a grid on the target image and mapping the grid to world coordinates, wherein the target image is captured by the camera; detecting the foot points of the first target based on deep learning and obtaining the image coordinates of the foot points in the target image; determining the first grid to which the foot point belongs based on the image coordinates of the foot point; mapping the image coordinates of the foot point to obtain the world coordinates of the foot point; and determining the true mapped position of the foot point according to the relationship between the first grid and the world coordinates of the foot point. This invention first uses automatic mapping technology to quickly obtain preliminary world coordinate mapping results, and then introduces a manual correction step to allow users to fine-tune key points according to actual conditions. This invention fully utilizes the speed advantage of automatic mapping, while improving the positioning accuracy of distant targets through manual correction, ensuring that the generated heatmap is smoother and more accurate. This invention uses the camera's intrinsic parameter matrix to perform distortion correction and back-projection processing on the image coordinates, converting them into normalized camera coordinates. Next, based on known planar information (such as the ground or a tabletop), a fixed depth value (z-coordinate) is set to determine the actual three-dimensional position of each target in the camera coordinate system. The camera's extrinsic parameters (rotation matrix and translation vector) are applied to convert these camera coordinates into world coordinates. This invention simplifies the complexity of depth estimation and can also efficiently generate heatmaps.

[0097] Based on the same inventive concept, a coordinate mapping heatmap generation device based on mesh optimization is provided, the device comprising:

[0098] The mapping module is used to draw a grid on the target image after the camera is calibrated, and to map the grid to world coordinates. The target image is captured by the camera.

[0099] The acquisition module is used to detect the foot points of the first target based on deep learning and obtain the image coordinates of the foot points in the target image.

[0100] The first grid module is used to determine the first grid to which a foot belongs based on its image coordinates.

[0101] The world coordinates module is used to map the image coordinates of the foot point to obtain the world coordinates of the foot point;

[0102] The mapping point module is used to determine the actual mapping point of the foot point based on the relationship between the world coordinates of the first grid and the foot point.

[0103] Based on the same inventive concept, the present invention also provides an electronic device, comprising:

[0104] processor;

[0105] Memory used to store processor-executable instructions;

[0106] The processor is configured to execute a grid-optimized coordinate mapping heatmap generation method as described above.

[0107] Based on the same inventive concept, the present invention also provides a non-transitory computer-readable storage medium, which, when the instructions in the storage medium are executed by the processor of an electronic device, enables the electronic device to execute the grid-optimized coordinate mapping heatmap generation method provided above.

[0108] Since the electronic device described in this embodiment is an electronic device used to implement the information processing method in the embodiments of the present invention, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the information processing method described in the embodiments of the present invention. Therefore, how the electronic device implements the method in the embodiments of the present invention will not be described in detail here. Any electronic device used by those skilled in the art to implement the information processing method in the embodiments of the present invention falls within the scope of protection of the present invention.

[0109] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0110] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0111] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0112] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0113] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0114] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for generating coordinate mapping heatmaps based on mesh optimization, characterized in that, The method includes: After the camera is calibrated, a grid is drawn on the target image and mapped to world coordinates, wherein the target image is captured by the camera; The foot points of the first target are detected based on deep learning, and the image coordinates of the foot points in the target image are obtained. Based on the image coordinates of the foot point, determine the first grid to which the foot point belongs; The image coordinates of the foot point are mapped to obtain the world coordinates of the foot point; Based on the relationship between the world coordinates of the first grid and the foot point, the actual mapped position of the foot point is determined.

2. The coordinate mapping heatmap generation method based on mesh optimization as described in claim 1, characterized in that, Based on the relationship between the first grid and the world coordinates of the footpoint, the true mapped position of the footpoint is determined, including: Determine whether the world coordinates of the foot point are within the world coordinates corresponding to the first grid. If the foot is in the world coordinates, then the true mapping point of the foot is the world coordinates of the foot; if the foot is not in the world coordinates, then a line is drawn connecting the center of the first grid and the foot in the world coordinates, and the intersection of the line and the first grid is the true mapping point of the foot.

3. The coordinate mapping heatmap generation method based on mesh optimization as described in claim 1, characterized in that, Also includes: Where P is the actual mapped position of the foot point, G is the first grid, and p aut Let G be the world coordinates of the foot point. edge The boundary of the first grid, p ma The coordinates of the foot point are shown in the image.

4. The coordinate mapping heatmap generation method based on mesh optimization as described in claim 1, characterized in that, Camera calibration includes: Where, k 3×3 R is the intrinsic parameter matrix of the camera. 3×3 Let t be a rotation matrix. 3×1 The vector is a translation vector, where u and v represent image coordinates, x, y, and h are world coordinates, and z is the translation vector. c This is for depth information.

5. The coordinate mapping heatmap generation method based on mesh optimization as described in claim 4, characterized in that, Unifying several objectives onto a single plane includes: Where, r 11 to r 33 All are elements in the rotation matrix, z is the depth parameter, t1 to t3 are translation vectors, and f x f y u0 and v0 are all elements in the intrinsic parameter matrix.

6. The coordinate mapping heatmap generation method based on mesh optimization as described in claim 1, characterized in that, Camera calibration also includes: Configure the intrinsic parameters of the camera, including: setting the focal length, principal point, or distortion coefficient of the camera; Configure the extrinsic parameters of the camera, including setting the rotation matrix and translation vector of the camera.

7. The coordinate mapping heatmap generation method based on mesh optimization as described in claim 1, characterized in that, The first target's footpoints are detected based on deep learning, including: The neural network to be trained is trained based on the YOLOv5 algorithm to obtain a human detection model. The human detection model is used to identify several targets in the target image and configure a followable bounding box for each target. The positions of several key points of the target are obtained based on a preset attitude estimation model; Extract the target's foot points from several key locations.

8. A coordinate mapping heatmap generation device based on mesh optimization, characterized in that, The device includes: The mapping module is used to draw a grid on the target image after the camera is calibrated, and to map the grid to world coordinates, wherein the target image is captured by the camera; The acquisition module is used to detect the foot points of the first target based on deep learning, and to acquire the image coordinates of the foot points in the target image. The first grid module is used to determine the first grid to which the foot point belongs based on the image coordinates of the foot point; The world coordinate module is used to map the image coordinates of the foot point to obtain the world coordinates of the foot point; The mapping point module is used to determine the actual mapping point of the foot point based on the relationship between the world coordinates of the first grid and the foot point.

9. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the grid-optimized coordinate mapping heatmap generation method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to perform the grid-optimized coordinate mapping heatmap generation method as described in any one of claims 1 to 7.

Images