Traffic accident scene digital reconstruction method based on ground-air cooperation

Abstract

The invention discloses a traffic accident scene digital reconstruction method based on ground-air cooperation. The method comprises the following steps: acquiring image data of an accident scene by using a five-point calibration method and a three-direction camera shooting method; performing denoising, enhancement and geometric distortion correction on the image data of the accident scene to obtain preprocessed image data; key feature points are extracted from the preprocessed image data, and feature point matching is carried out between different images to obtain a feature point matching result; and reconstructing an accident scene digital reconstruction model according to a feature point matching result and the preprocessed image data. According to the invention, the precision of a measurement result can be ensured, and the safety of traffic accident scene investigation work and the comprehensiveness of investigation data are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of accident scene reconstruction, and in particular to a method for digitally reconstructing a traffic accident scene based on ground-air collaboration. Background Art

[0002] According to statistics and analysis by the World Health Organization (WHO), traffic accidents are seriously affecting human health and safety. The property losses caused by traffic accidents have also led to serious social and economic burdens. Efficient and orderly handling of traffic accidents can not only reflect the city's management capabilities, but also alleviate traffic congestion and avoid secondary accidents. Traffic accident site investigation is an important procedure for accident handling. The relevant data obtained through the investigation is the main basis for restoring the accident process and determining the responsibility for the accident. How to complete the traffic accident investigation work comprehensively, accurately and efficiently and provide a traceable technical path is the challenge faced at this stage.

[0003] The current means of traffic accident investigation mainly relies on the on-site hand-drawn schematic diagrams of front-line police officers. First, the baseline and benchmark points of the accident scene are calibrated, and then the key locations on the scene are marked. Tools such as tape measures are used to measure and record the distances between the key locations and the baseline and benchmark points to complete the accident scene investigation.

[0004] (1) Due to the different working experiences of police officers, the investigation is often incomplete and the data is incomplete, which brings great difficulties to the reconstruction of the accident process and the determination of facts.

[0005] (2) The use of traditional survey equipment such as tape measures cannot be standardized, and errors are difficult to control when measuring large distances.

[0006] (3) Most of the on-site traces are formed on the road surface and will disappear after the accident. This makes it impossible to trace the on-site measurement data, making on-site resurvey and accident reconstruction difficult.

[0007] (4) Traditional survey and measurement rely on on-site measurement by personnel using measuring tapes, which is inefficient and can easily lead to traffic congestion and even secondary accidents. Summary of the invention

[0008] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0009] To this end, the present invention proposes a method for digital reconstruction of traffic accident scenes based on ground-air collaboration. Through ground-air collaboration, unmanned aerial vehicle camera technology, accident road surface precise calibration technology, and a reconstruction algorithm model based on ground-air parameters are integrated to develop an intelligent digital reconstruction model of traffic accident scenes. Corresponding solutions are proposed to the problems of comprehensiveness, accuracy and efficiency of traffic accident scene investigation.

[0010] Another object of the present invention is to propose a traffic accident scene digital reconstruction system based on ground-air collaboration.

[0011] To achieve the above-mentioned purpose, the present invention proposes a method for digitally reconstructing a traffic accident scene based on ground-air collaboration, comprising:

[0012] The image data of the accident scene is obtained by using the five-point calibration method and the three-position camera method;

[0013] De-noising, enhancing and geometric distortion correction are performed on the image data of the accident scene to obtain pre-processed image data;

[0014] Extract key feature points from preprocessed image data, and perform feature point matching between different images to obtain feature point matching results;

[0015] A digital reconstruction model of the accident scene is reconstructed according to the feature point matching results and the pre-processed image data.

[0016] The method for digitally reconstructing a traffic accident scene based on ground-air collaboration in an embodiment of the present invention may also have the following additional technical features:

[0017] In one embodiment of the present invention, after reconstructing the three-dimensional digital model of the accident scene according to the feature point matching results and the pre-processed image data, the method further includes:

[0018] Input the marking information of any two points on the accident road surface into the accident scene digital reconstruction model to output the distance between the two points.

[0019] In one embodiment of the present invention, the five-point calibration method comprises:

[0020] Determine the scope that needs to be calibrated based on the size and shape of the accident scene;

[0021] Place a set of reflective rods at the northeast, southeast, southwest and northwest corners of the accident scene to determine the spatial scope and boundaries of the accident scene; each set of reflective rods contains two rods, which are placed perpendicular to each other;

[0022] Place a set of reflective poles in the middle of the accident scene;

[0023] The five sets of reflective rod lengths were used as measurement references to calculate the reference distance of the accident scene.

[0024] In one embodiment of the present invention, the three-position camera method includes:

[0025] Use a drone to take photos from three different angles of the accident scene to obtain three images; each image contains ground details and spatial relationships from different perspectives of the accident scene;

[0026] Obtain three-dimensional information of ground objects by performing stereoscopic vision measurement on three images;

[0027] Based on the three-dimensional information, the feature points in the three images are matched, and the matched feature points are used for image stitching to form a complete ground image map.

[0028] In one embodiment of the present invention, a drone is used to photograph the accident scene from three different angles to obtain three images, including:

[0029] Use a drone to take three-dimensional photos at the accident scene, including vertical overhead photos at the middle position and 60-degree overhead photos at two boundary positions, to obtain multi-angle images; during the shooting process, record the internal and external parameters of the drone;

[0030] The recorded intrinsic parameters are used to perform geometric correction on the acquired multi-angle images.

[0031] To achieve the above-mentioned purpose, a second embodiment of the present invention proposes a traffic accident scene digital reconstruction system based on ground-air collaboration, comprising:

[0032] An image data acquisition module is used to acquire image data of the accident scene using a five-point calibration method and a three-position camera method;

[0033] A data preprocessing module is used to perform denoising, enhancement and geometric distortion correction on the image data of the accident scene to obtain preprocessed image data;

[0034] A feature point matching module is used to extract key feature points from preprocessed image data and perform feature point matching between different images to obtain feature point matching results;

[0035] The digital reconstruction module is used to reconstruct a digital reconstruction model of the accident scene according to the feature point matching results and the pre-processed image data.

[0036] The method and system for digital reconstruction of traffic accident scenes based on ground-air collaboration of the embodiments of the present invention establish a digital model of the accident scene by coupling multiple links of technology such as accident road surface calibration, drone photography, and PC software reconstruction, and complete the traffic accident scene investigation work efficiently, accurately, and comprehensively.

[0037] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0039] Figure 1 is an architecture diagram of a method for digitally reconstructing a traffic accident scene based on ground-air collaboration according to an embodiment of the present invention;

[0040] Figure 2 is a flow chart of a method for digitally reconstructing a traffic accident scene based on ground-air collaboration according to an embodiment of the present invention;

[0041] Figure 3 is a schematic diagram of a five-point three-orientation method of a traffic accident scene digital reconstruction platform according to an embodiment of the present invention;

[0042] Figure 4 It is a structural diagram of a traffic accident scene digital reconstruction system based on ground-air collaboration according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0044] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0045] The following describes a method and system for digital reconstruction of a traffic accident scene based on ground-air collaboration according to an embodiment of the present invention with reference to the accompanying drawings.

[0046] Currently, in the research on traffic accident scene investigation, there is a lack of intelligent measurement tools based on digital reconstruction platforms. The present invention has carried out a system design on the ground-air collaborative photogrammetry technology. Utilizing the existing photogrammetry technology, a digital reconstruction platform of the "five-point three-direction method" is designed and constructed. The digital reconstruction platform established by this invention can be based on an efficient method of ground markings and drone photography. After obtaining the internal and external parameters of the drone, it can perform digital reconstruction on the PC side and output the distance between any two points on the road surface at the scene of the accident without any safety risks. The present invention is mainly used in traffic accident scene investigation work, and can also be extended to field investigation measurement and digital reconstruction in other fields. The overall architecture is as follows: Figure 1 shown.

[0047] Figure 2 is a flow chart of a method for digitally reconstructing a traffic accident scene based on ground-air collaboration according to an embodiment of the present invention. Figure 2 As shown, including:

[0048] S1, using the five-point calibration method and three-position camera method to obtain image data of the accident scene.

[0049] Specifically, the accident road surface is calibrated using the "five-point calibration method", that is, 1-3m reflective poles are selected according to the size of the accident scene and placed at the four azimuths and the middle position of the accident scene. The five-point calibration method completes the calibration of the spatial position boundary and the selection of the reference distance of the accident scene.

[0050] It is understandable that the five-point calibration method is a commonly used method to restore the geometric relationship of a three-dimensional scene in a two-dimensional image. It is suitable for estimating the scope of the accident scene. The five-point calibration method is mainly used for camera calibration. The intrinsic parameters of the camera, such as the focal length of the lens, wide-angle characteristics, etc., and the spatial posture of the camera relative to the scene are determined by selecting five calibration points.

[0051] It can be understood that the general five-point calibration method includes 1. Selecting the appropriate ruler size: Selecting the appropriate ruler size to ensure that each ruler point can be clearly distinguished in the photos taken at the accident scene. The ruler size should be large enough to occupy a certain proportion in the photo, but not too large to interfere with other important information at the scene. 2. Placing the ruler at five points: Select five points at the accident scene to place the ruler. These five points should be distributed at different locations at the accident scene and try to cover the entire area of ​​interest to ensure the accuracy of the calibration. The placement of the ruler should be stable to prevent movement due to wind or other external forces. 3. Shooting height and overhead angle: When shooting, the appropriate height and overhead angle should be selected so that the camera can look down on the entire accident scene from above. The purpose of this is to reduce the distortion caused by perspective and keep the image plane consistency as much as possible. 4. Lens focal length and lens wide angle: Select the appropriate lens focal length and wide angle characteristics. Wide-angle lenses may introduce more edge distortion, so special attention should be paid when shooting. Standard focal length lenses usually provide a more natural perspective and reduce distortion. 5. Parameters such as image distortion: During post-processing, the image needs to be corrected based on the lens parameters used during shooting, such as focal length, wide-angle characteristics, etc., to eliminate or reduce image distortion.

[0052] In one embodiment of the present invention, Figure 3 As shown, the detailed calibration process of the present invention is as follows:

[0053] Determine the scope of the accident scene: Determine the scope that needs to be calibrated based on the size and shape of the accident scene.

[0054] Choose reflective poles: Choose reflective poles that are 1-5m long, and make sure they are long enough to be clearly identified in the drone footage.

[0055] Placement of reflective rods:

[0056] Four angles: Place a group of reflective rods at the northeast, southeast, southwest, and northwest angles of the accident scene. Each group of reflective rods contains two rods placed perpendicular to each other.

[0057] Center location: Place a set of reflective rods in the center of the accident scene.

[0058] Marking the ground: This is an optional step. If necessary, you can use markers or colors to mark the position of the reflective rods on the ground for subsequent shooting and measurement.

[0059] Therefore, the five-point calibration method of the present invention has the following technical effects:

[0060] Spatial position boundary calibration: The spatial scope and boundaries of the accident scene can be determined by reflective rods at four azimuth angles.

[0061] Reference distance selection: Five sets of reflective rod lengths are used as measurement references to calculate the distance between any two points at the accident scene.

[0062] Subsequent data traceability: The coordinates of the reflective rod can be used as a benchmark for subsequent data reconstruction and measurement to ensure the accuracy and traceability of the data.

[0063] Simple and easy: easy to operate, no complicated equipment and technology required.

[0064] Efficient and convenient: The spatial position boundary calibration and reference distance selection of the accident scene can be completed quickly.

[0065] Data accuracy: Provides an accurate benchmark for subsequent data reconstruction and measurement to ensure data accuracy and reliability.

[0066] In one embodiment of the present invention, the drone photography is mainly based on the principles of digital orthophoto and stereo vision measurement, using the "three-position photography method": taking vertical overhead photos at the middle position and 60-degree overhead photos at the boundary positions of two adjacent positions, to obtain three-position drone digital images. Figure 3 shown.

[0067] Specifically, the three-position camera method includes the following steps:

[0068] Step 1: Accident scene preparation

[0069] Pre-flight: This step aims to verify the on-site scope confirmed by the “five-point calibration method” and determine the aerial photography altitude and scale length.

[0070] Drone parameters: Set and record the internal and external parameters of the drone, including focal length, wide angle, lens distortion, etc., as well as the drone's flight altitude and overhead shooting angle.

[0071] Step 2: Drone photography

[0072] Take a vertical aerial photo of the middle position of the accident scene. The drone flies above the middle position of the accident scene and takes a photo vertically downward to ensure that the photo covers the entire accident scene area and the outline of the ruler is clearly visible.

[0073] Take a 60-degree aerial photo at two adjacent azimuths. The drone flies above two adjacent azimuths (such as due north and due west) and takes a 60-degree aerial photo at the ground, taking one photo each time. Ensure that the two photos can cover the entire accident scene area. The images should have overlapping parts and can reflect the positions and contours of the five scales at the scene.

[0074] S2, denoising, enhancing and geometric distortion correction are performed on the image data of the accident scene to obtain pre-processed image data.

[0075] S3, extracting key feature points from the preprocessed image data, and performing feature point matching between different images to obtain feature point matching results.

[0076] S4, reconstructing the digital reconstruction model of the accident scene based on the feature point matching results and the pre-processed image data.

[0077] Specifically, the present invention uses PC software to establish an accident scene data model based on the above-mentioned "five-point three-direction method" camera results. The model should include image data after removing terrain deformation, lens distortion, etc.; at the same time, it should be able to obtain three-dimensional information of road objects through images in different directions to obtain more accurate data measurement results.

[0078] Application of the five-point calibration method of three-azimuth photography in the reconstruction process of the present invention: The five-point calibration method establishes an accurate coordinate reference by placing reflective rods at four azimuths and the middle position of the accident scene. The position and spacing of these reflective rods can be determined, thereby providing a reliable reference for subsequent image data processing and three-dimensional reconstruction. Due to the distortion of the drone camera lens, there is a deviation between the captured image and the actual scene. The software algorithm will correct the image according to the lens parameters and the distortion model, eliminate the error caused by the distortion, and ensure the accuracy of subsequent measurements. The five-point calibration method not only provides a ground coordinate reference, but also realizes the accurate matching of ground data and aerial images by fusing with drone image data, further improving the measurement accuracy.

[0079] The application of three-dimensional photography in the reconstruction process of the present invention: three-dimensional photography captures the accident scene from different angles to obtain rich information. These images contain ground details and spatial relationships from different perspectives, providing a sufficient data basis for subsequent three-dimensional reconstruction. By performing stereoscopic visual measurement on the three images, three-dimensional information of ground objects can be obtained, including height, depth, etc. This enables the platform to perform more accurate distance measurements, such as the distance between a vehicle and the ground, and between vehicles. The software algorithm will match and splice the feature points in the image, and merge the three images into a complete ground image map. After correction and splicing, this image map can more realistically reflect the situation at the accident scene and provide a reliable data source for subsequent measurement and analysis.

[0080] Therefore, the present invention performs denoising, enhancement, correction and other operations to improve the image quality and provide a better data basis for subsequent processing. Feature points are extracted from the image and matched to establish the corresponding relationship between the images, providing a basis for subsequent three-dimensional reconstruction and measurement. According to the image data and the matching results of the feature points, the three-dimensional digital model of the accident scene is reconstructed using triangulation and Bundle Adjustment technology. Based on the three-dimensional model, any distance in the reconstructed model can be measured as needed, such as the distance between the vehicle and the ground, between vehicles, the length of road surface marks, the size of road facilities, etc.

[0081] It can be understood that the present invention establishes a digital link relationship of road surface dimensions under ground-air coordination based on the three-dimensional information of the scene and the digital orthophoto. Figure 1 shown.

[0082] It can be understood that based on the above-mentioned link relationship, the accuracy of size measurement can be optimized by changing the shooting angle (three different directions) while fixing a ruler. The specific steps are as follows: Ruler positioning: Place a ruler of known size in the target area. Multi-angle shooting: Shoot the target from three different directions. Image processing: Use image processing technology to extract the location information of the ruler from each picture. Data fusion: Combine the data of the three pictures and optimize the size measurement results through mathematical operations.

[0083] It can be understood that based on the above-mentioned link relationship, under the condition of fixed shooting orientation, five rulers are used to improve the accuracy of dimension measurement, including ruler layout: evenly arrange five rulers of known sizes in the target area. Shooting: shoot the entire scene from a fixed angle. Feature extraction: identify and extract the position information of all rulers from the image. Operation optimization: based on the position information of the five rulers, the accuracy of dimension measurement is optimized through mathematical operation methods.

[0084] Furthermore, the marking information of any two points on the accident road surface is input into the accident scene digital reconstruction model to output the distance between the two points.

[0085] It is understandable that the reconstructed accident scene data model can save the data set based on the length of the ground markings and meet the intelligent calculation of the distance between any two points on the road surface in the image, which provides technical conditions for tracing the source of the accident scene data and on-site re-survey. In special cases, if three-dimensional measurement of vertical distance is required, the vertical reference size needs to be provided at the scale stage.

[0086] The digital reconstruction method of traffic accident scene based on ground-air collaboration according to the embodiment of the present invention has intelligent and efficient data output capability, and uses digital orthophoto and stereo vision measurement to ensure the accuracy of the measurement results. It can ensure the safety of on-site investigation and the comprehensiveness of investigation data. The invention can be suitable for investigation and measurement of traffic accidents of different scopes, can be applicable to small-scale and small-vehicle traffic accidents on urban roads, and can also complete work in complex and dangerous accident scene investigation environments such as expressways and highways.

[0087] In order to implement the above embodiment, Figure 4 As shown, this embodiment also provides a traffic accident scene digital reconstruction system 10 based on ground-air collaboration, including:

[0088] The image data acquisition module 100 is used to acquire the image data of the accident scene by using the five-point calibration method and the three-position camera method;

[0089] The data preprocessing module 200 is used to perform denoising, enhancement and geometric distortion correction on the image data of the accident scene to obtain preprocessed image data;

[0090] The feature point matching module 300 is used to extract key feature points from the pre-processed image data and perform feature point matching between different images to obtain feature point matching results;

[0091] The digital reconstruction module 400 is used to reconstruct the digital reconstruction model of the accident scene according to the feature point matching results and the pre-processed image data.

[0092] In one embodiment of the present invention, after the digital reconstruction module 400, the following is further included:

[0093] The calculation output module is used to input the marking information of any two points on the accident road surface into the accident scene digital reconstruction model to output the distance between the two points.

[0094] In one embodiment of the present invention, the image data acquisition module 100 is further used for:

[0095] Determine the scope that needs to be calibrated based on the size and shape of the accident scene;

[0096] Place a set of reflective rods at the northeast, southeast, southwest and northwest corners of the accident scene to determine the spatial scope and boundaries of the accident scene; each set of reflective rods contains two rods, which are placed perpendicular to each other;

[0097] Place a set of reflective poles in the middle of the accident scene;

[0098] The five sets of reflective rod lengths were used as measurement references to calculate the reference distance of the accident scene.

[0099] In one embodiment of the present invention, the image data acquisition module 100 is further used for:

[0100] Use a drone to take photos from three different angles of the accident scene to obtain three images; each image contains ground details and spatial relationships from different perspectives of the accident scene;

[0101] Obtain three-dimensional information of ground objects by performing stereoscopic vision measurement on three images;

[0102] Based on the three-dimensional information, the feature points in the three images are matched, and the matched feature points are used for image stitching to form a complete ground image map.

[0103] In one embodiment of the present invention, a drone is used to photograph the accident scene from three different angles to obtain three images, including:

[0104] Use a drone to take three-dimensional photos at the accident scene, including vertical overhead photos at the middle position and 60-degree overhead photos at two boundary positions, to obtain multi-angle images; during the shooting process, record the internal and external parameters of the drone;

[0105] The recorded intrinsic parameters are used to perform geometric correction on the acquired multi-angle images.

[0106] The digital reconstruction system of the traffic accident scene based on ground-air collaboration according to the embodiment of the present invention has intelligent and efficient data output capabilities, and uses digital orthophoto and stereo vision measurement to ensure the accuracy of the measurement results. It can ensure the safety of on-site investigation work and the comprehensiveness of investigation data. The invention can be suitable for the investigation and measurement of traffic accidents of different scopes, can be applicable to small-scale and small-vehicle traffic accidents on urban roads, and can also complete work in complex and dangerous accident scene investigation environments such as expressways and highways.

[0107] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0108] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

Claims

1. A method for digital reconstruction of traffic accident scene based on ground-air collaboration, characterized in that: include: The image data of the accident scene is obtained by using the five-point calibration method and the three-position camera method; De-noising, enhancing and geometric distortion correction are performed on the image data of the accident scene to obtain pre-processed image data; Extract key feature points from preprocessed image data, and perform feature point matching between different images to obtain feature point matching results; A digital reconstruction model of the accident scene is reconstructed according to the feature point matching results and the pre-processed image data.

2. The method according to claim 1, characterized in that After reconstructing the three-dimensional digital model of the accident scene according to the feature point matching results and the pre-processed image data, the method further includes: Input the marking information of any two points on the accident road surface into the accident scene digital reconstruction model to output the distance between the two points.

3. The method according to claim 1, characterized in that The five-point calibration method comprises: Determine the scope that needs to be calibrated based on the size and shape of the accident scene; Place a set of reflective rods at the northeast, southeast, southwest and northwest corners of the accident scene to determine the spatial scope and boundaries of the accident scene; each set of reflective rods contains two rods, which are placed perpendicular to each other; Place a set of reflective poles in the middle of the accident scene; The five sets of reflective rod lengths were used as measurement references to calculate the reference distance of the accident scene.

4. The method according to claim 1, characterized in that: The three-position camera method comprises: Use a drone to take photos from three different angles of the accident scene to obtain three images; each image contains ground details and spatial relationships from different perspectives of the accident scene; Obtain three-dimensional information of ground objects by performing stereoscopic vision measurement on three images; Based on the three-dimensional information, the feature points in the three images are matched, and the matched feature points are used for image stitching to form a complete ground image map.

5. The method according to claim 4, characterized in that The drone was used to take photos from three different angles of the accident scene to obtain three images, including: Use a drone to take three-dimensional photos at the accident scene, including vertical overhead photos at the middle position and 60-degree overhead photos at two boundary positions, to obtain multi-angle images; during the shooting process, record the internal and external parameters of the drone; The recorded intrinsic parameters are used to perform geometric correction on the acquired multi-angle images.

6. A traffic accident scene digital reconstruction system based on ground-air collaboration, characterized in that: include: An image data acquisition module is used to acquire image data of the accident scene using a five-point calibration method and a three-position camera method; A data preprocessing module is used to perform denoising, enhancement and geometric distortion correction on the image data of the accident scene to obtain preprocessed image data; A feature point matching module is used to extract key feature points from preprocessed image data and perform feature point matching between different images to obtain feature point matching results; The digital reconstruction module is used to reconstruct a digital reconstruction model of the accident scene according to the feature point matching results and the pre-processed image data.

7. The system according to claim 6, characterized in that After the digital reconstruction module, it also includes: The calculation output module is used to input the marking information of any two points on the accident road surface into the accident scene digital reconstruction model to output the distance between the two points.

8. The system according to claim 6, characterized in that The image data acquisition module is also used for: Determine the scope that needs to be calibrated based on the size and shape of the accident scene; Place a set of reflective rods at the northeast, southeast, southwest and northwest corners of the accident scene to determine the spatial scope and boundaries of the accident scene; each set of reflective rods contains two rods, which are placed perpendicular to each other; Place a set of reflective poles in the middle of the accident scene; The five sets of reflective rod lengths were used as measurement references to calculate the reference distance of the accident scene.

9. The system according to claim 6, characterized in that The image data acquisition module is also used for: Use a drone to take photos from three different angles of the accident scene to obtain three images; each image contains ground details and spatial relationships from different perspectives of the accident scene; Obtain three-dimensional information of ground objects by performing stereoscopic vision measurement on three images; Based on the three-dimensional information, the feature points in the three images are matched, and the matched feature points are used for image stitching to form a complete ground image map.

10. The system according to claim 9, characterized in that The drone was used to take photos from three different angles of the accident scene to obtain three images, including: Use a drone to take three-dimensional photos at the accident scene, including vertical overhead photos at the middle position and 60-degree overhead photos at two boundary positions, to obtain multi-angle images; during the shooting process, record the internal and external parameters of the drone; The recorded intrinsic parameters are used to perform geometric correction on the acquired multi-angle images.

Images