Road traffic accident three-dimensional reconstruction system and method

Abstract

The invention relates to the technical field of auxiliary processing of road traffic accidents, in particular to a three-dimensional reconstruction system and method for road traffic accidents, and the system comprises a data collection module, a scene modeling module and an accident reconstruction module. The data acquisition module is used for acquiring environment data and evidence data of a target road traffic accident scene; the scene modeling module is used for performing three-dimensional scene modeling based on the environmental data to obtain a three-dimensional scene model of the target road traffic accident scene; and the accident reconstruction module is used for performing accident process reconstruction based on the evidence data and the three-dimensional scene model, and obtaining traffic participant operation data and three-dimensional animation of the target road traffic accident scene. According to the invention, rapid and accurate data acquisition can be realized, an accident scene can be constructed and reconstructed, and an accident process can be scientifically and visually restored and displayed.

Description

Technical Field

[0001] The present invention relates to the technical field of auxiliary processing of road traffic accidents, and in particular to a three-dimensional reconstruction system and method for road traffic accidents. Background Art

[0002] Road traffic accident reconstruction is of great significance for restoring the accident process, analyzing the causes of the accident, and serving accident prevention. Currently, accident reconstruction data relies on accident scene investigation, and accident scene investigation is mainly carried out by measuring with a tape measure, photographic recording, and on-site marking. This recording method is not only time-consuming and labor-intensive and prone to traffic congestion, but also prone to errors in data collection. The risk exposure of investigators is high, and re-investigation cannot be achieved after evacuating the scene. For complex accidents that lack relevant evidence, there are problems such as lack of basis for accident process reconstruction, non-intuitiveness, and high reproduction costs. Therefore, the present invention provides a three-dimensional reconstruction system and method for road traffic accidents. Summary of the Invention

[0003] The purpose of the present invention is to provide a three-dimensional reconstruction system and method for road traffic accidents to address the problems existing in current accident investigations, such as low efficiency of road traffic accident scene investigations, incomplete evidence preservation, and unclear accident process, so as to achieve fast and accurate data collection and construct a reconstructed accident scene, and scientifically and intuitively restore and display the accident process.

[0004] To achieve the above-mentioned object, the present invention provides a road traffic accident three-dimensional reconstruction system, comprising: a data acquisition module, a scene modeling module and an accident reconstruction module;

[0005] The data acquisition module is used to collect environmental data and evidence data at the target road traffic accident scene;

[0006] The scene modeling module is used to perform three-dimensional scene modeling based on the environmental data to obtain a three-dimensional scene model of the target road traffic accident scene;

[0007] The accident reconstruction module is used to reconstruct the accident process based on the evidence data and the three-dimensional scene model, and obtain the traffic participant operation data and three-dimensional animation at the target road traffic accident scene.

[0008] Optionally, the data acquisition module includes an acquisition unit and a control unit, wherein the acquisition unit is used to acquire image data, three-dimensional point cloud data and evidence data of the target road traffic accident scene; and the control unit is used to control the data acquisition work of the acquisition unit.

[0009] Optionally, the scene modeling module includes a data acquisition unit, a three-dimensional modeling unit and a data export unit, wherein the data acquisition unit is used to acquire the image data and three-dimensional point cloud data collected by the data acquisition module in real time; the three-dimensional modeling unit is used to perform three-dimensional modeling based on the image data and three-dimensional point cloud data to obtain the three-dimensional scene model; and the data export unit is used to export the three-dimensional scene model.

[0010] Optionally, the accident reconstruction module includes an environment reconstruction unit, a trajectory reconstruction unit and a visualization unit, wherein the environment reconstruction unit is used to obtain a static road environment model by assigning dynamic properties to the three-dimensional scene model; the trajectory reconstruction unit is used to simulate the running trajectories of traffic participants before, during and after a traffic accident based on the evidence data; the visualization unit is used to fuse the running trajectories and the static road environment model, and output the running data of each traffic participant and a three-dimensional animation of the accident process.

[0011] On the other hand, the present invention also provides a method for 3D reconstruction of a road traffic accident, comprising:

[0012] Develop a data collection plan based on the conditions of the target road traffic accident scene;

[0013] Controlling a data acquisition module to acquire environmental data and evidence data at a target road traffic accident scene based on the data acquisition scheme, wherein the environmental data includes three-dimensional point cloud data and image data;

[0014] Inputting the environmental data into a scene modeling module for three-dimensional modeling to obtain a three-dimensional scene model;

[0015] The evidence data and the three-dimensional scene model are input into an accident reconstruction module to reconstruct the accident process, and the running data and three-dimensional animation of the traffic participants at the target road traffic accident scene are obtained.

[0016] Optionally, reconstructing the accident process includes:

[0017] Assigning dynamic properties to the three-dimensional scene model to obtain a static road environment model;

[0018] Simulating the running trajectories of each traffic participant at the target road traffic accident scene before, during, and after the traffic accident based on the evidence data;

[0019] The running trajectory and the static road environment model are integrated to output the running data of each traffic participant and a three-dimensional animation of the accident process.

[0020] Optionally, the three-dimensional scene model includes a point cloud data model and an image data model, wherein the point cloud data model is obtained by correcting and trimming the three-dimensional point cloud data and then modeling it; the image data model is used to enhance and denoise the image data and then model it.

[0021] Optionally, assigning dynamic attributes to the three-dimensional scene model includes:

[0022] triangulating the point cloud data model into a three-dimensional surface and assigning the dynamic properties;

[0023] The image data model is used as a 3D road object and is endowed with the dynamic attributes.

[0024] The beneficial effects of the present invention are:

[0025] The present invention can realize convenient and fast data collection through the acquisition equipment, improve the exploration efficiency and reduce the time that the investigators are exposed to dangerous environments; through the three-dimensional modeling of point cloud data and image data, the error can reach the centimeter level and the millimeter level, and the exploration accuracy is higher than manual measurement; and three-dimensional modeling technology, as an efficient and accurate three-dimensional data acquisition and processing method, can quickly and accurately collect and fix the accident scene information, reconstruct the accident scene, and after the accident scene is evacuated, if there is a need for re-survey, it can be quickly and accurately surveyed in the three-dimensional modeling scene to make up for the problems such as omission of early information; compared with reconnaissance experiments under real conditions, the accident process is simulated by computer simulation, which not only provides guidance for case detection and responsibility determination, but also greatly reduces the investigation cost, and can realize fast and accurate collection and reconstruction of accident scenes, and scientifically and intuitively restore and display the accident process. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. 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 any creative work.

[0027] Figure 1 This is a schematic structural diagram of a road traffic accident 3D reconstruction system according to an embodiment of the present invention;

[0028] Figure 2 This is a flow chart of a three-dimensional reconstruction method for a road traffic accident according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0030] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] On the one hand, this embodiment provides a road traffic accident 3D reconstruction system, such as Figure 1 As shown, it includes: data acquisition module, scenario modeling module and accident reconstruction module;

[0032] The data acquisition module is used to collect environmental data and evidence data at the target road traffic accident scene;

[0033] The scene modeling module is used to perform three-dimensional scene modeling based on the environmental data to obtain a three-dimensional scene model of the target road traffic accident scene;

[0034] The accident reconstruction module is used to reconstruct the accident process based on the evidence data and the three-dimensional scene model, and obtain the traffic participant operation data and three-dimensional animation at the target road traffic accident scene.

[0035] Specifically, this embodiment can realize convenient and rapid data collection through the acquisition equipment, thereby improving the efficiency of investigation and reducing the time that investigators are exposed to dangerous environments; three-dimensional modeling is performed through point cloud data and image data, and the error can reach the centimeter and millimeter levels, and the investigation accuracy is higher than manual measurement; and three-dimensional modeling technology, as an efficient and accurate three-dimensional data acquisition and processing method, can quickly and accurately collect and fix accident scene information and reconstruct the accident scene. After evacuating the accident scene, if there is a need for re-survey, it can be quickly and accurately surveyed in the three-dimensional modeling scene to make up for problems such as omissions of early information; compared with reconnaissance experiments under real conditions, the simulation of the accident process through computer simulation not only provides guidance for case detection and responsibility determination, but also greatly reduces the investigation cost, and can realize rapid and accurate collection and reconstruction of accident scenes, scientifically and intuitively restore and display the accident process.

[0036] Furthermore, the data acquisition module includes an acquisition unit and a control unit, wherein the acquisition unit is used to acquire image data, three-dimensional point cloud data and evidence data of the target road traffic accident scene; and the control unit is used to control the data acquisition work of the acquisition unit.

[0037] Specifically, in this embodiment, the acquisition unit and control unit, respectively, employ a data acquisition device and a controller for collecting 3D scene data from road traffic accidents. The data acquisition device, equipped with sensors such as cameras and lidar, is used to perceive and collect information about the accident scene. Examples include handheld 3D laser scanners, drones equipped with point cloud laser scanners, or cameras. The controller controls the data acquisition device and provides real-time feedback on the data collected.

[0038] Furthermore, the scene modeling module includes a data acquisition unit, a three-dimensional modeling unit and a data export unit, wherein the data acquisition unit is used to acquire the image data and three-dimensional point cloud data collected by the data acquisition module in real time; the three-dimensional modeling unit is used to perform three-dimensional modeling based on the image data and three-dimensional point cloud data to obtain the three-dimensional scene model; and the data export unit is used to export the three-dimensional scene model.

[0039] Specifically, in this embodiment, the scene modeling module models, processes, and converts the data collected by the data acquisition module into a data format file compatible with the accident reconstruction unit. The data acquisition unit receives and stores data collected by the data acquisition device. The 3D modeling unit uses the collected data to generate a 3D scene model of the traffic accident, including a point cloud data model and an image data model (drone oblique photography model). The data export unit exports the 3D scene model into a 3D model file in a fixed format. Point cloud data model formats include E57, XYZ, and LAS, and image data model formats include 3DS, OBJ, DAE, STL, WRL, KMZ, and LWO.

[0040] Furthermore, the accident reconstruction module includes an environment reconstruction unit, a trajectory reconstruction unit and a visualization unit, wherein the environment reconstruction unit is used to obtain a static road environment model by assigning dynamic properties to the three-dimensional scene model; the trajectory reconstruction unit is used to simulate the running trajectories of traffic participants before, during and after a traffic accident based on the evidence data; the visualization unit is used to fuse the running trajectories with the static road environment model, and output the running data of each traffic participant and a three-dimensional animation of the accident process.

[0041] Specifically, the accident reconstruction module in this embodiment uses existing evidence to restore the static road environment and dynamic running trajectory of the traffic accident, including:

[0042] First, the 3D traffic accident scene model is imported and processed through the environmental reconstruction unit to give it dynamic properties. Specifically, for the point cloud data model, the accident area, such as the road surface and collision objects, is triangulated into a 3D surface model and given dynamic properties. The main operation process is as follows:

[0043] (1) Import point cloud data.

[0044] (2) Segmentation of point cloud. To facilitate triangulation of various areas (such as roads, elevations, ditches, etc.), the point cloud is first segmented into sections, and new layers are created for each section as needed.

[0045] (3) Triangulation: Define the area of the point cloud as the slope polygon to be considered in the simulation. First, set the view to only show the layers that need to be triangulated. Next, define the minimum and maximum triangulation distances. It is recommended to set the maximum triangulation distance to 5m to avoid generating overly large triangulation triangles. Finally, define the selected area as the slope polygon.

[0046] For oblique photography models, dynamic properties can be given by directly importing them as 3D road objects.

[0047] Then, the trajectory reconstruction unit is used to simulate and restore the running trajectory of each traffic participant before, during, and after the accident. The kinematic theory is followed before the accident, and the dynamic theory is followed after the accident. The process is as follows:

[0048] (1) Load the accident vehicle data, check and modify the vehicle parameters, and select the appropriate model.

[0049] (2) Based on the environmental reconstruction results, the collision point of the vehicles is determined, and the relative positions of the two vehicles embedded in each other are set according to the deformation conditions.

[0050] (3) Analyze and determine the sequence of events in the accident process.

[0051] (4) Set the vehicle's driving path.

[0052] (5) Analyze and set the intermediate position and stop position of the vehicle in the event of an accident.

[0053] (6) Simulate, analyze and optimize the accident collision process.

[0054] Finally, the visualization unit is used to output the reconstructed traffic participant operation data and display three-dimensional animation to complete the accident reconstruction.

[0055] The units in the scenario modeling module and the accident reconstruction module in this embodiment run on a computer, enabling data flow within the computer. The data acquisition device and the controller are connected via wireless transmission methods such as a local area network and Bluetooth, while the data acquisition device and the data acquisition unit are connected via wired transmission.

[0056] On the other hand, this embodiment also provides a method for 3D reconstruction of a road traffic accident, including:

[0057] Develop a data collection plan based on the conditions of the target road traffic accident scene;

[0058] Controlling a data acquisition module to acquire environmental data and evidence data at a target road traffic accident scene based on the data acquisition scheme, wherein the environmental data includes three-dimensional point cloud data and image data;

[0059] Inputting the environmental data into a scene modeling module for three-dimensional modeling to obtain a three-dimensional scene model;

[0060] The evidence data and the three-dimensional scene model are input into an accident reconstruction module to reconstruct the accident process, and the running data and three-dimensional animation of the traffic participants at the target road traffic accident scene are obtained.

[0061] Furthermore, reconstructing the accident process includes:

[0062] Assigning dynamic properties to the three-dimensional scene model to obtain a static road environment model;

[0063] Simulating the running trajectories of each traffic participant at the target road traffic accident scene before, during, and after the traffic accident based on the evidence data;

[0064] The running trajectory and the static road environment model are integrated to output the running data of each traffic participant and a three-dimensional animation of the accident process.

[0065] Furthermore, the three-dimensional scene model includes a point cloud data model and an image data model, wherein the point cloud data model is obtained by correcting and trimming the three-dimensional point cloud data and then modeling it; the image data model is used to enhance and denoise the image data and then model it.

[0066] Furthermore, assigning dynamic attributes to the three-dimensional scene model includes:

[0067] triangulating the point cloud data model into a three-dimensional surface and assigning the dynamic properties;

[0068] The image data model is used as a 3D road object and is endowed with the dynamic attributes.

[0069] The following combination Figure 2 By way of example, the specific workflow of the 3D reconstruction method for a road traffic accident proposed in this embodiment is described as follows:

[0070] (1) 3D reconstruction of road traffic accidents based on point cloud data.

[0071] (1) Preparation stage: First, select a 3D scanning device equipped with laser radar as the data collection device, such as a handheld 3D scanner. After assembling the handheld 3D scanner, connect the corresponding software using the local area network and Bluetooth, debug and ensure that the device is in normal operation. Secondly, define the data collection scope of the accident scene: According to the situation of the traffic accident scene, the data collection scope of the accident scene is clearly defined as 50 meters in front of the direction of the vehicle or the location of other participants 5 seconds before the accident to the position where the vehicle stopped after the collision, including the collision point, and ensure that the collection scope includes the road traffic environment of the entire road section / intersection and the relevant circumstances of the accident. Then, plan the data collection route: The principles to be followed when planning the route are: ① The route should be continuously and evenly distributed within the collection range, not too dense or too sparse; ② The periphery of the route should cover the collection range; ③ Try to use serpentine movement to ensure the quality of the collected point cloud; ④ Try to choose a scanning route with a wide field of view, such as the middle of the road; ⑤ The route should be as non-repetitive as possible; ⑥ The route can be adjusted according to the real-time situation of point cloud collection.

[0072] (2) Data collection phase: Collect data according to the collection scope and planned route in the data collection plan. Check the collection status on the software during collection and adjust the collection plan appropriately when necessary to ensure that the road environment point cloud record is complete until data collection is completed.

[0073] (3) Scenario modeling: First, connect the data acquisition device and the computer, import the collected field data into the computer via wired transmission, and store it in a management folder. Second, use point cloud data processing software to correct and trim the point cloud data and focus on the accident section. Finally, export the processed point cloud data model into E57, XYZ, or LAS format.

[0074] (4) Accident reconstruction phase: Use PC-Crash software to complete the accident reconstruction. First, import the point cloud data model into PC-Crash. Second, segment the point cloud representing the road and environment into different layers, triangulate the road layer point cloud into slope polygons, so that it has dynamic properties and participates in the simulation as an object to complete the road environment reconstruction. Third, using existing evidence, implement the accident collision simulation based on dynamics and kinematics in PC-Crash to complete the participant trajectory reconstruction. Finally, output the traffic participant operation data and 3D animation.

[0075] (2) Three-dimensional reconstruction of road traffic accidents based on drone oblique photography.

[0076] (1) Preparation stage: First, select a drone equipped with a camera as the data acquisition device, use the remote control as the controller to connect wirelessly to the drone, use the local area network and Bluetooth to connect to the corresponding software, debug and ensure that the equipment is in normal operation. Secondly, define the scope of data collection at the accident scene: Based on the situation at the traffic accident scene, the data collection scope at the accident scene is clearly defined as 50 meters in front of the direction of the vehicle or the location of other participants 5 seconds before the accident to the position where the vehicle stopped after the collision, including the collision point, to ensure that the collection range includes the road traffic environment of the entire road section / intersection and the relevant circumstances involved in the accident. Then, plan the data collection route: When planning the route, the planning object is the drone flight route, and the drone route planning principles are followed.

[0077] (2) Data collection phase: Collect data according to the collection scope and planned route in the data collection plan. Check the collection status on the software during collection and adjust the collection plan appropriately when necessary to ensure that the road environment point cloud record is complete until data collection is completed.

[0078] (3) Scene modeling stage: First, connect the data acquisition equipment and the computer, import the collected field data into the computer via wired transmission, and store it in a management folder. Second, process the image data collected by the drone aerial photography to generate an oblique photography model, and export the model into 3DS, OBJ, DAE, STL, WRL, KMZ, and LWO format files.

[0079] (4) Accident reconstruction: Accident reconstruction is completed using PC-Crash software. First, the oblique photography model is imported into PC-Crash as a 3D road object, giving it dynamic properties and completing the road environment reconstruction. Second, using existing evidence, a dynamic and kinematic accident collision simulation is implemented in PC-Crash to complete the participant trajectory reconstruction. Finally, the traffic participant movement data and 3D animation are output.

[0080] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A road traffic accident 3D reconstruction system, characterized in that: include: Data acquisition module, scenario modeling module and accident reconstruction module; The data acquisition module is used to collect environmental data and evidence data at the target road traffic accident scene; The scene modeling module is used to perform three-dimensional scene modeling based on the environmental data to obtain a three-dimensional scene model of the target road traffic accident scene; The accident reconstruction module is used to reconstruct the accident process based on the evidence data and the three-dimensional scene model, and obtain the traffic participant operation data and three-dimensional animation at the target road traffic accident scene.

2. The road traffic accident 3D reconstruction system according to claim 1, characterized in that: The data acquisition module includes an acquisition unit and a control unit, wherein the acquisition unit is used to acquire image data, three-dimensional point cloud data and evidence data of the target road traffic accident scene; and the control unit is used to control the data acquisition work of the acquisition unit.

3. The road traffic accident 3D reconstruction system according to claim 2, characterized in that: The scene modeling module includes a data acquisition unit, a three-dimensional modeling unit and a data export unit, wherein the data acquisition unit is used to acquire the image data and three-dimensional point cloud data collected by the data acquisition module in real time; the three-dimensional modeling unit is used to perform three-dimensional modeling based on the image data and three-dimensional point cloud data to obtain the three-dimensional scene model; and the data export unit is used to export the three-dimensional scene model.

4. The road traffic accident 3D reconstruction system according to claim 3, characterized in that: The accident reconstruction module includes an environment reconstruction unit, a trajectory reconstruction unit and a visualization unit, wherein the environment reconstruction unit is used to obtain a static road environment model by assigning dynamic properties to the three-dimensional scene model; the trajectory reconstruction unit is used to simulate the running trajectories of traffic participants before, during and after a traffic accident based on the evidence data; the visualization unit is used to fuse the running trajectories with the static road environment model and output the running data of each traffic participant and a three-dimensional animation of the accident process.

5. A method for 3D reconstruction of a road traffic accident, implemented based on the 3D reconstruction system for road traffic accidents according to any one of claims 1 to 4, characterized in that: include: Develop a data collection plan based on the conditions of the target road traffic accident scene; Controlling a data acquisition module to acquire environmental data and evidence data at a target road traffic accident scene based on the data acquisition scheme, wherein the environmental data includes three-dimensional point cloud data and image data; Inputting the environmental data into a scene modeling module for three-dimensional modeling to obtain a three-dimensional scene model; The evidence data and the three-dimensional scene model are input into an accident reconstruction module to reconstruct the accident process, and the running data and three-dimensional animation of the traffic participants at the target road traffic accident scene are obtained.

6. The three-dimensional reconstruction method of a road traffic accident according to claim 5, characterized in that: Reconstructing the accident process includes: Assigning dynamic properties to the three-dimensional scene model to obtain a static road environment model; Simulating the running trajectories of each traffic participant at the target road traffic accident scene before, during, and after the traffic accident based on the evidence data; The running trajectory and the static road environment model are integrated to output the running data of each traffic participant and a three-dimensional animation of the accident process.

7. The method for 3D reconstruction of a road traffic accident according to claim 6, characterized in that: The three-dimensional scene model includes a point cloud data model and an image data model, wherein the point cloud data model is obtained by correcting and trimming the three-dimensional point cloud data and then modeling it; the image data model is used to enhance and denoise the image data and then model it.

8. The method for 3D reconstruction of a road traffic accident according to claim 7, characterized in that: Assigning dynamic attributes to the three-dimensional scene model includes: The point cloud data model is triangulated into a three-dimensional surface and endowed with the dynamic properties; the image data model is used as a 3D road object and endowed with the dynamic properties.

Images