Automatic road traffic accident scene map making method based on three-dimensional laser scanning technology

Abstract

The invention discloses an automatic road traffic accident scene map making method based on a three-dimensional laser scanning technology, relates to the technical field of traffic accident processing, and solves the problem of lack of a three-dimensional laser scanner for quickly acquiring three-dimensional space information of an accident scene, generating dense point cloud data and providing a data basis for accident analysis. The technical problems that in the prior art, a high-precision three-dimensional laser scanning technology and an automatic data processing method are difficult to combine, and whole-course automatic processing from data acquisition to field map generation is difficult to achieve are solved. Encrypting the data; data transmission; integrity detection; carrying out point cloud color assignment; an orthoimage is overlooked; automatically identifying a candidate area; automatically drawing a road contour; optimizing the contour line; outputting a result; and encrypting the result. The traffic accident scene map can be quickly generated, and the accident investigation efficiency and quality are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of traffic accident handling, and in particular relates to an automated road traffic accident scene map production method based on three-dimensional laser scanning technology. Background Art

[0002] With the development of intelligent transportation systems, rapid and accurate investigation of traffic accidents has become a key link in improving traffic management efficiency and accident handling speed. Therefore, an automated road traffic accident scene map production method based on 3D laser scanning technology has emerged. By applying advanced 3D laser scanning technology and automated data processing methods, rapid and accurate generation of traffic accident scene maps can be achieved, thereby improving the efficiency and quality of accident investigation.

[0003] Although the existing traffic accident scene investigation has achieved traffic accident investigation to a certain extent, it relies on manual measurement and hand-drawn drawings, and has problems such as long time consumption, low precision, and susceptibility to human factors. There is a lack of 3D laser scanners to quickly obtain 3D spatial information of the accident scene and generate dense point cloud data to provide a data basis for accident analysis. It is difficult to combine high-precision 3D laser scanning technology and automated data processing methods, and it is also difficult to achieve full automation from data collection to scene map generation. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an automated road traffic accident scene map production method based on three-dimensional laser scanning technology, which is used to solve the following technical problems:

[0005] Although the existing traffic accident scene investigation has achieved traffic accident investigation to a certain extent, it relies on manual measurement and hand-drawn drawings, and has problems such as long time consumption, low precision, and susceptibility to human factors. There is a lack of 3D laser scanners to quickly obtain 3D spatial information of the accident scene and generate dense point cloud data to provide a data basis for accident analysis. It is difficult to combine high-precision 3D laser scanning technology and automated data processing methods, and it is also difficult to achieve full automation from data collection to scene map generation.

[0006] To solve the above problems, the first aspect of the present invention provides an automated road traffic accident scene map production method based on three-dimensional laser scanning technology, comprising the following steps:

[0007] S1: Collect point cloud data and image data of the accident scene;

[0008] S2: Use AES encryption algorithm to encrypt the collected data;

[0009] S3: Transmit the encrypted data to a remote server or a designated computer or tablet device;

[0010] S4: During the transmission process, CRC check is used to implement data integrity check;

[0011] S5: Accurately match the point cloud data with the high-definition image, use image fusion technology to assign the matched color information to the point cloud data, and generate a color point cloud;

[0012] S6: Convert the color point cloud data into a bird's-eye view to generate a high-definition bird's-eye orthophoto;

[0013] S7: Using the intensity information of the point cloud data, the candidate road boundary areas are identified through the threshold segmentation algorithm;

[0014] S8: Use the RANSAC algorithm to automatically extend the line segments to form the road contour. For curved line segments, use the least squares method to fit the circle to make them straight or broken lines.

[0015] S9: Perform vertical projection based on the part above the ground point in the point cloud data to generate a preliminary outline of the traffic participation system, and use morphological processing to optimize the outline;

[0016] S10: Combining the road lines and the outlines of traffic participants, the system automatically draws a two-dimensional vector map of the traffic accident scene and outputs orthophotos and drawing files;

[0017] S11: Use digital signature technology to encrypt the generated report, use a randomly generated key, and send it to relevant personnel through a secure channel.

[0018] As a further solution of the present invention, step S1 comprises the following steps:

[0019] In the data collection area, a comprehensive scanning device integrating 3D laser scanner, high-definition infrared camera, GPS, gyroscope and accelerometer sensors is used to conduct a comprehensive scan of the road traffic accident scene; the laser scanner emits laser beams and receives reflected light, calculates distance and angle, and generates dense point cloud data; the built-in GPS, gyroscope and accelerometer record the scanning position, scanning direction and instantaneous acceleration in real time, and mark the position space and time; the integrated high-definition infrared camera synchronously takes photos of the scene.

[0020] As a further solution of the present invention, the AES encryption algorithm comprises the following steps:

[0021] S201: Each byte of each plaintext block is replaced by another byte;

[0022] S202: The rows of the state matrix are shifted according to the rules;

[0023] S203: treating each column of the state matrix as a polynomial and performing multiplication operation on a fixed polynomial;

[0024] S204: At the beginning and end of each round, the state matrix is ​​XORed with a round key.

[0025] As a further solution of the present invention, the transmission mode selects wireless transmission or wired transmission according to the on-site network environment.

[0026] As a further solution of the present invention, the CRC check comprises the following steps:

[0027] S401: Setting a generating polynomial G(x);

[0028] S402: Expand the polynomial F(x) corresponding to the data to be sent, use G(x) as the divisor, perform modulo 2 division on the expanded polynomial, and the remainder obtained is the r-bit CRC check code;

[0029] S403: append the check code to the end of the original data to form a new data sequence and send it to the receiving end;

[0030] S404: After receiving the data, the receiving end uses G(x) to perform a modulo 2 division operation. If the remainder obtained is 0, it means that there is no error in the data transmission process; if it is not 0, it means that the data is erroneous.

[0031] As a further solution of the present invention, the threshold segmentation algorithm comprises the following steps:

[0032] S701: Convert the input color image into a grayscale image;

[0033] S702: Filtering the grayscale image;

[0034] S703: selecting a threshold value according to the characteristics of the image;

[0035] S704: Compare the grayscale value of each pixel in the preprocessed grayscale image with the selected threshold value. If the grayscale value of the pixel is higher than the threshold value, mark the pixel as a road area; if the grayscale value of the pixel is not higher than the threshold value, mark it as a background area; and segment the image into two areas, the road area and the background area.

[0036] S705: performing morphological operations on the segmented image;

[0037] S706: Detect the position of the road boundary by using a Canny edge detector.

[0038] As a further solution of the present invention, the S703 includes the following steps:

[0039] The threshold is selected according to the characteristics of the image. There are two situations for selecting the threshold;

[0040] If the contrast between the target and the background in the image is relatively uniform, a single threshold is used for segmentation of the entire image;

[0041] If the contrast between the target and background in the image is uneven, use local thresholding or adaptive thresholding to dynamically adjust the threshold according to the characteristics of different areas in the image.

[0042] As a further solution of the present invention, the S706 includes the following steps:

[0043] S7061: Perform convolution operation on the image using Gaussian filter;

[0044] Gaussian filter formula:

[0045]

[0046] Where f(x,y) is a two-dimensional Gaussian function, x and y are the coordinates of a position in the template relative to the center of the template, σ is the standard deviation of the Gaussian function, and π is a constant in the calculation formula;

[0047] S7062: Calculate the gradient of the image in the horizontal and vertical directions using a first-order derivative operator;

[0048] Gradient strength and gradient direction calculation formula:

[0049]

[0050] Among them, G is the gradient strength, θ is the gradient direction, G x and G y are the gradients of the image in the horizontal and vertical directions respectively;

[0051] S7063: For each pixel, compare its gradient strength with the gradient strength of the adjacent pixel in the gradient direction. If the gradient strength of the pixel is not the largest, suppress it and set it to 0; if it is the largest, retain it.

[0052] S7064: Set two thresholds: a low threshold and a high threshold;

[0053] Pixels with gradient strength greater than the high threshold are identified as strong edge points and retained directly;

[0054] Pixels with gradient strength less than the low threshold are identified as non-edge points and directly suppressed;

[0055] Pixels with gradient strength between the two thresholds are identified as potential edge points. Check whether there are strong edge points in their neighborhood. If there are, they are retained as edge points, otherwise, they are suppressed.

[0056] S7065: By iteratively visiting potential edge points and checking strong edge points in their neighborhoods, potential edge points are connected with strong edge points to form a complete edge line.

[0057] As a further solution of the present invention, the RANSAC algorithm comprises the following steps:

[0058] S801: Randomly select two points from the data set as the minimum sample set;

[0059] S802: Calculate the parameters of the straight line equation using the two selected points;

[0060] S803: Substitute each point in the data set into the straight line equation, calculate the distance from the point to the straight line, and according to the set distance threshold, regard the point whose distance is less than the threshold as an inner point, otherwise regard it as an outer point;

[0061] S804: Count the number of inliers of the current model. If the number of inliers of the current model is greater than the number of inliers of the best model previously recorded, update the best model and its inlier set.

[0062] S805: Repeat steps S801-S804 until a preset number of iterations is reached or the number of inliers of the best model no longer increases significantly;

[0063] S806: Output the optimal model parameters and the corresponding inlier point set.

[0064] Beneficial effects of the present invention:

[0065] The present invention can quickly acquire three-dimensional spatial information of the accident scene through a three-dimensional laser scanner, generate dense point cloud data, and provide a rich data basis for accident analysis. The transmitted data is encrypted through the AES encryption algorithm to provide high-intensity security protection. The CRC check is used for integrity verification to achieve error detection and data integrity verification, and improve the reliability of data transmission. The intensity information of the point cloud data is used to identify the candidate road boundary area through a threshold segmentation algorithm to achieve accuracy and efficiency in road feature extraction. In combination with high-precision three-dimensional laser scanning technology and automated data processing algorithms, full automated processing from data acquisition to scene map generation is achieved, providing a new technical means for the rapid and accurate handling of traffic accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0067] Figure 1 It is a flowchart of the implementation process of the present invention. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 are within the scope of protection of the present invention.

[0069] See also Figure 1 As shown, the present invention is an automated road traffic accident scene map production method based on three-dimensional laser scanning technology, comprising the following steps:

[0070] S1: Collect point cloud data and image data of the accident scene;

[0071] S2: Use AES encryption algorithm to encrypt the collected data;

[0072] S3: Transmit the encrypted data to a remote server or a designated computer or tablet device;

[0073] S4: During the transmission process, CRC check is used to implement data integrity check;

[0074] S5: Accurately match the point cloud data with the high-definition image, use image fusion technology to assign the matched color information to the point cloud data, and generate a color point cloud;

[0075] S6: Convert the color point cloud data into a bird's-eye view to generate a high-definition bird's-eye orthophoto;

[0076] S7: Using the intensity information of the point cloud data, the candidate road boundary areas are identified through the threshold segmentation algorithm;

[0077] S8: Use the RANSAC algorithm to automatically extend the line segments to form the road contour. For curved line segments, use the least squares method to fit the circle to make them straight or broken lines.

[0078] S9: Perform vertical projection based on the part above the ground point in the point cloud data to generate a preliminary outline of the traffic participation system, and use morphological processing to optimize the outline;

[0079] S10: Combining the road lines and the outlines of traffic participants, the system automatically draws a two-dimensional vector map of the traffic accident scene and outputs orthophotos and drawing files;

[0080] S11: Use digital signature technology to encrypt the generated report, use a randomly generated key, and send it to relevant personnel through a secure channel.

[0081] Specifically, data is collected through integrated scanning equipment, encrypted using the AES encryption algorithm, and the encrypted data is transmitted to a remote device. CRC check is used for integrity detection, and image fusion technology is used to color the point cloud to generate a high-definition bird's-eye view orthophoto. The threshold segmentation algorithm is used to identify candidate road boundary areas, and the RANSAC algorithm is used to automatically draw the road contour to generate a preliminary contour of the traffic participation system. The system automatically draws a two-dimensional vector traffic accident scene map, which is encrypted using digital signature technology.

[0082] In one embodiment of the present invention, the step S1 comprises the following steps:

[0083] In the data collection area, a comprehensive scanning device integrating 3D laser scanner, high-definition infrared camera, GPS, gyroscope and accelerometer sensors is used to conduct a comprehensive scan of the road traffic accident scene; the laser scanner emits laser beams and receives reflected light, calculates distance and angle, and generates dense point cloud data; the built-in GPS, gyroscope and accelerometer record the scanning position, scanning direction and instantaneous acceleration in real time, and mark the position space and time; the integrated high-definition infrared camera synchronously takes photos of the scene.

[0084] Specifically, the laser scanner emits a laser beam and receives reflected light, calculates distance and angle, and generates dense point cloud data. The built-in GPS, gyroscope, and accelerometer record the scanning position including latitude, longitude, and altitude, the scanning direction including angle and rotation speed, and instantaneous acceleration in real time, providing accurate spatial and temporal markings for the data. The integrated high-definition infrared camera takes photos of the scene synchronously to ensure that each key area has a clear image record.

[0085] In one embodiment of the present invention, the AES encryption algorithm comprises the following steps:

[0086] S201: Each byte of each plaintext block is replaced by another byte;

[0087] S202: The rows of the state matrix are shifted according to the rules;

[0088] S203: treating each column of the state matrix as a polynomial and performing multiplication operation on a fixed polynomial;

[0089] S204: At the beginning and end of each round, the state matrix is ​​XORed with a round key.

[0090] Specifically, each plaintext block is usually 128 bits, i.e., 16 bytes, where the replacement is done based on a fixed, predefined S-box table, which maps input bytes to output bytes; the rows of the state matrix are shifted according to the rules, and the state matrix is ​​a 4*4 byte matrix, where the first row is not shifted, the second row is shifted left by one bit, the third row is shifted left by two bits, and the fourth row is shifted left by three bits; each column of the state matrix is ​​regarded as a polynomial, and after multiplication with a fixed polynomial, a polynomial over GF(2^8), the result is modulo an irreducible polynomial, usually x^8+x^4+x^3+x+1, to ensure that the result is still an 8-bit byte; the state matrix is ​​XORed with a round key, where the round key is generated by the original key through a key expansion algorithm.

[0091] In one embodiment of the present invention, the transmission mode is selected as wireless transmission or wired transmission according to the on-site network environment.

[0092] Specifically, wireless transmission includes 4G, 5G and WI-FI, or limited transmission includes network cable, type-C to USB and type-C data cable.

[0093] In one embodiment of the present invention, the CRC check comprises the following steps:

[0094] S401: Setting a generating polynomial G(x);

[0095] S402: Expand the polynomial F(x) corresponding to the data to be sent, use G(x) as the divisor, perform modulo 2 division on the expanded polynomial, and the remainder obtained is the r-bit CRC check code;

[0096] S403: append the check code to the end of the original data to form a new data sequence and send it to the receiving end;

[0097] S404: After receiving the data, the receiving end uses G(x) to perform a modulo 2 division operation. If the remainder obtained is 0, it means that there is no error in the data transmission process; if it is not 0, it means that the data is erroneous.

[0098] Specifically, a generating polynomial G(x) is set, and the format of the polynomial is pre-set by the sending and receiving ends; the polynomial F(x) corresponding to the data to be sent is expanded, and the data to be sent is set to m bits. The expansion is usually left shifted by r bits. G(x) is used as the divisor, and a modulo 2 division operation is performed on the expanded polynomial. The remainder obtained is the r-bit CRC check code; the check code is appended to the end of the original data to form a new data sequence and send it to the receiving end; after the receiving end receives the data, G(x) is also used to perform a modulo 2 division operation. If the remainder obtained is 0, it means that there is no error in the data transmission process; if it is not 0, it means that the data is incorrect.

[0099] In one embodiment of the present invention, the threshold segmentation algorithm comprises the following steps:

[0100] S701: Convert the input color image into a grayscale image;

[0101] S702: Filtering the grayscale image;

[0102] S703: selecting a threshold value according to the characteristics of the image;

[0103] S704: Compare the grayscale value of each pixel in the preprocessed grayscale image with the selected threshold value. If the grayscale value of the pixel is higher than the threshold value, mark the pixel as a road area; if the grayscale value of the pixel is not higher than the threshold value, mark it as a background area; and segment the image into two areas, the road area and the background area.

[0104] S705: performing morphological operations on the segmented image;

[0105] S706: Detect the position of the road boundary by using a Canny edge detector.

[0106] Specifically, the input color image is converted into a grayscale image. The grayscale image can simplify calculations and retain the brightness information in the image, which is crucial for the recognition of road boundaries. The grayscale image is filtered, such as using a median filter or a Gaussian filter, to remove noise and unnecessary details in the image and improve the image quality. A suitable threshold is selected, and the selection of the threshold depends on the characteristics of the image. The grayscale value of each pixel in the preprocessed grayscale image is compared with the selected threshold. When the grayscale value of the pixel is higher than the threshold, the pixel is marked as a road area; if it is lower than the threshold, it is marked as a background area, and the image can be segmented into two areas, the road and the background, so as to extract the candidate area of ​​the road boundary. Morphological operations such as corrosion, expansion, opening and closing operations are performed on the segmented image to further smooth the road boundary, remove small noise points, and fill the holes in the road area. The Canny edge detector is used to detect the precise position of the road boundary to improve the accuracy of recognition.

[0107] In one embodiment of the present invention, the S703 includes the following steps:

[0108] The threshold is selected according to the characteristics of the image. There are two situations for selecting the threshold;

[0109] If the contrast between the target and the background in the image is relatively uniform, a single threshold is used for segmentation of the entire image;

[0110] If the contrast between the target and background in the image is uneven, use local thresholding or adaptive thresholding to dynamically adjust the threshold according to the characteristics of different areas in the image.

[0111] Specifically, there are two situations for selecting the threshold; when the contrast between the target and the background in the image is relatively uniform, a single threshold is used to segment the entire image; when the contrast between the target and the background in the image is uneven, a local threshold or an adaptive threshold is used to dynamically adjust the threshold according to the characteristics of different areas in the image.

[0112] In one embodiment of the present invention, the S706 includes the following steps:

[0113] S7061: Perform convolution operation on the image using Gaussian filter;

[0114] Gaussian filter formula:

[0115]

[0116] Where f(x,y) is a two-dimensional Gaussian function, x and y are the coordinates of a position in the template relative to the center of the template, σ is the standard deviation of the Gaussian function, and π is a constant in the calculation formula;

[0117] S7062: Calculate the gradient of the image in the horizontal and vertical directions using a first-order derivative operator;

[0118] Gradient strength and gradient direction calculation formula:

[0119]

[0120] Among them, G is the gradient strength, θ is the gradient direction, G x and G y are the gradients of the image in the horizontal and vertical directions respectively;

[0121] S7063: For each pixel, compare its gradient strength with the gradient strength of the adjacent pixel in the gradient direction. If the gradient strength of the pixel is not the largest, suppress it and set it to 0; if it is the largest, retain it.

[0122] S7064: Set two thresholds: a low threshold and a high threshold;

[0123] Pixels with gradient strength greater than the high threshold are identified as strong edge points and retained directly;

[0124] Pixels with gradient strength less than the low threshold are identified as non-edge points and directly suppressed;

[0125] Pixels with gradient strength between the two thresholds are identified as potential edge points. Check whether there are strong edge points in their neighborhood. If there are, they are retained as edge points, otherwise, they are suppressed.

[0126] S7065: By iteratively visiting potential edge points and checking strong edge points in their neighborhoods, potential edge points are connected with strong edge points to form a complete edge line.

[0127] Specifically, a Gaussian filter is used to perform a convolution operation on the image. The Gaussian filter is a linear smoothing filter whose weight is determined by a Gaussian function. The filter can effectively blur the image while retaining edge information. A first-order derivative operator is used to calculate the gradient of the image in the horizontal and vertical directions. For each pixel, its gradient strength is compared with the gradient strength of the adjacent pixel points in its gradient direction. If the gradient strength of the pixel point is not the largest, it is suppressed and set to 0, otherwise it is retained. Since the gradient direction is continuous, the gradient direction is usually divided into several discrete directions such as horizontal, vertical, and two diagonal directions in actual implementation, and comparison is made in these directions. Two thresholds are set, including a low threshold and a high threshold, and the type of edge point is determined based on the comparison of the gradient strength of the pixel point with the two thresholds. By iteratively accessing potential edge points and checking the strong edge points in their neighborhood, the potential edge points are connected with the strong edge points to form a complete edge line.

[0128] In one embodiment of the present invention, the RANSAC algorithm comprises the following steps:

[0129] S801: Randomly select two points from the data set as the minimum sample set;

[0130] S802: Calculate the parameters of the straight line equation using the two selected points;

[0131] S803: Substitute each point in the data set into the straight line equation, calculate the distance from the point to the straight line, and according to the set distance threshold, regard the point whose distance is less than the threshold as an inner point, otherwise regard it as an outer point;

[0132] S804: Count the number of inliers of the current model. If the number of inliers of the current model is greater than the number of inliers of the best model previously recorded, update the best model and its inlier set.

[0133] S805: Repeat steps S801-S804 until a preset number of iterations is reached or the number of inliers of the best model no longer increases significantly;

[0134] S806: Output the optimal model parameters and the corresponding inlier point set.

[0135] Specifically, two points are randomly selected from the data set as the minimum sample set; two points are selected to calculate the parameters of the line equation, such as the slope k and the intercept b, or other line representation methods such as the normal vector and a point are used; each point in the data set is substituted into the line equation, its distance to the line is calculated, and whether it is an inlier is determined according to the set distance threshold; the number of inliers of the current model is counted, and if the number of inliers of the current model is greater than the number of inliers of the best model previously recorded, the best model and its inlier set are updated; steps S801-S804 are repeated until the preset number of iterations is reached or the number of inliers of the best model no longer increases significantly; the best model parameters and the corresponding inlier set are output. The inliers can be used to extend the line segment, for example, the inliers are fitted by the least squares method to obtain the final line segment equation to ensure the continuity and non-overlap of the road lines.

[0136] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology, characterized in that: The following steps are involved: S1: Collect point cloud data and image data of the accident scene; S2: Use AES encryption algorithm to encrypt the collected data; S3: Transmit the encrypted data to a remote server or a designated computer or tablet device; S4: During the transmission process, CRC check is used to implement data integrity check; S5: Accurately match the point cloud data with the high-definition image, use image fusion technology to assign the matched color information to the point cloud data, and generate a color point cloud; S6: Convert the color point cloud data into a bird's-eye view to generate a high-definition bird's-eye orthophoto; S7: Using the intensity information of the point cloud data, the candidate road boundary areas are identified through the threshold segmentation algorithm; S8: Use the RANSAC algorithm to automatically extend the line segments to form the road contour. For curved line segments, use the least squares method to fit the circle to make them straight or broken lines. S9: Perform vertical projection based on the part above the ground point in the point cloud data to generate a preliminary outline of the traffic participation system, and use morphological processing to optimize the outline; S10: Combining the road lines and the outlines of traffic participants, the system automatically draws a two-dimensional vector map of the traffic accident scene and outputs orthophotos and drawing files; S11: Use digital signature technology to encrypt the generated report, use a randomly generated key, and send it to relevant personnel through a secure channel.

2. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 1, characterized in that: The step S1 comprises the following steps: In the data collection area, a comprehensive scanning device integrating 3D laser scanner, high-definition infrared camera, GPS, gyroscope and accelerometer sensors is used to conduct a comprehensive scan of the road traffic accident scene; the laser scanner emits laser beams and receives reflected light, calculates distance and angle, and generates dense point cloud data; the built-in GPS, gyroscope and accelerometer record the scanning position, scanning direction and instantaneous acceleration in real time, and mark the position space and time; the integrated high-definition infrared camera synchronously takes photos of the scene.

3. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 1, characterized in that: The AES encryption algorithm comprises the following steps: S201: Each byte of each plaintext block is replaced by another byte; S202: The rows of the state matrix are shifted according to the rules; S203: treating each column of the state matrix as a polynomial and performing multiplication operation on a fixed polynomial; S204: At the beginning and end of each round, the state matrix is ​​XORed with a round key.

4. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 1, characterized in that: The transmission mode is selected as wireless transmission or wired transmission according to the on-site network environment.

5. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 1, characterized in that: The CRC check comprises the following steps: S401: Setting a generating polynomial G(x); S402: Expand the polynomial F(x) corresponding to the data to be sent, use G(x) as the divisor, perform modulo 2 division on the expanded polynomial, and the remainder obtained is the r-bit CRC check code; S403: append the check code to the end of the original data to form a new data sequence and send it to the receiving end; S404: After receiving the data, the receiving end uses G(x) to perform a modulo 2 division operation. If the remainder obtained is 0, it means that there is no error in the data transmission process; if it is not 0, it means that the data is erroneous.

6. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 1, characterized in that: The threshold segmentation algorithm comprises the following steps: S701: Convert the input color image into a grayscale image; S702: Filtering the grayscale image; S703: selecting a threshold value according to the characteristics of the image; S704: Compare the grayscale value of each pixel in the preprocessed grayscale image with the selected threshold value. If the grayscale value of the pixel is higher than the threshold value, mark the pixel as a road area; if the grayscale value of the pixel is not higher than the threshold value, mark it as a background area; and segment the image into two areas, the road area and the background area. S705: performing morphological operations on the segmented image; S706: Detect the position of the road boundary by using a Canny edge detector.

7. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 6, characterized in that: The S703 includes the following steps: The threshold is selected according to the characteristics of the image. There are two situations for selecting the threshold; If the contrast between the target and the background in the image is relatively uniform, a single threshold is used for segmentation of the entire image; If the contrast between the target and background in the image is uneven, use local thresholding or adaptive thresholding to dynamically adjust the threshold according to the characteristics of different areas in the image.

8. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 6, characterized in that: The S706 includes the following steps: S7061: Perform convolution operation on the image using Gaussian filter; Gaussian filter formula: Where f(x,y) is a two-dimensional Gaussian function, x and y are the coordinates of a position in the template relative to the center of the template, σ is the standard deviation of the Gaussian function, and π is a constant in the calculation formula; S7062: Calculate the gradient of the image in the horizontal and vertical directions using a first-order derivative operator; Gradient strength and gradient direction calculation formula: Among them, G is the gradient strength, θ is the gradient direction, G x and G y are the gradients of the image in the horizontal and vertical directions respectively; S7063: For each pixel, compare its gradient strength with the gradient strength of the adjacent pixel in the gradient direction. If the gradient strength of the pixel is not the largest, suppress it and set it to 0; if it is the largest, retain it. S7064: Set two thresholds: a low threshold and a high threshold; Pixels with gradient strength greater than the high threshold are identified as strong edge points and retained directly; Pixels with gradient strength less than the low threshold are identified as non-edge points and directly suppressed; Pixels with gradient strength between the two thresholds are identified as potential edge points. Check whether there are strong edge points in their neighborhood. If there are, they are retained as edge points, otherwise, they are suppressed. S7065: By iteratively visiting potential edge points and checking strong edge points in their neighborhoods, potential edge points are connected with strong edge points to form a complete edge line.

9. The method for producing an automated road traffic accident scene map based on three-dimensional laser scanning technology according to claim 1, characterized in that: The RANSAC algorithm comprises the following steps: S801: Randomly select two points from the data set as the minimum sample set; S802: Calculate the parameters of the straight line equation using the two selected points; S803: Substitute each point in the data set into the straight line equation, calculate the distance from the point to the straight line, and according to the set distance threshold, regard the point whose distance is less than the threshold as an inner point, otherwise regard it as an outer point; S804: Count the number of inliers of the current model. If the number of inliers of the current model is greater than the number of inliers of the best model previously recorded, update the best model and its inlier set. S805: Repeat steps S801-S804 until a preset number of iterations is reached or the number of inliers of the best model no longer increases significantly; S806: Output the optimal model parameters and the corresponding inlier point set.

Images