A method, apparatus, electronic device, and storage medium for accident reconstruction

Abstract

This application provides a method, apparatus, electronic device, and storage medium for accident reconstruction. The method includes: determining the flight path parameters of an aerial photography device based on a test image sequence of a test accident acquired by the device in a test scenario; acquiring a first image sequence captured by the aerial photography device during flight according to the flight path parameters in an actual scenario; wherein the first image sequence includes multiple frames of first images, with each subsequent frame being captured after optimizing the shooting parameters of the previous frame; and displaying a first accident scene reconstructed and simulated based on the first image sequence. This application, through efficient data acquisition by unmanned aerial vehicles and precise analysis using computer vision technology, can efficiently complete the three-dimensional reconstruction and dynamic analysis of accident scenes, providing a scientific basis for accident liability determination.

Description

Technical Field

[0001] This application relates to the field of accident reconstruction technology, and more specifically, to an accident reconstruction method, apparatus, electronic device, and storage medium. Background Technology

[0002] The primary task of accident reconstruction is to use the driving traces left on the road at the accident scene, the collision marks on the vehicle, and the vehicle's stopping position to deduce the vehicle's speed, acceleration, heading angle, and trajectory after the collision, in order to infer the cause of the traffic accident, determine the responsibility for the accident, and take appropriate action.

[0003] Traditional accident scene investigation mainly relies on manual measurement and photographic evidence collection. This method has problems such as insufficient accuracy of manual operation and long time consumption. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a method, apparatus, electronic device and storage medium for accident reconstruction to overcome the problems in the prior art.

[0005] In a first aspect, embodiments of this application provide a method for accident reconstruction, the method comprising:

[0006] In the test scenario, the flight path parameters of the aerial photography equipment are determined based on the test image sequence of the test accident acquired by the aerial photography equipment.

[0007] In a real-world scenario, the aerial photography equipment acquires a first image sequence during flight according to the flight path parameters; wherein, the first image sequence includes multiple first images, and each subsequent first image is obtained by optimizing the shooting parameters of the previous first image.

[0008] The first accident scene is shown in the reconstruction simulation based on the first image sequence.

[0009] In some technical solutions of this application, the flight path parameters of the aerial photography equipment are determined based on the test image sequence of the test accident acquired by the aerial photography equipment in the test scenario, including:

[0010] Based on the image overlap and path length between adjacent frames in the test image sequence, a bi-objective optimization model for the test image sequence is established.

[0011] By solving the bi-objective optimization model, the flight path parameters of the aerial photography equipment are obtained.

[0012] In some technical solutions of this application, the above-mentioned establishment of a bi-objective optimization model for the test image sequence based on the image overlap and path length between adjacent frames in the test image sequence includes:

[0013] Based on the original overlap rate, image quality factor, and terrain factor between adjacent frames in the test image sequence, an overlap evaluation sub-model is established.

[0014] Based on the distance between adjacent frames in the test image sequence and the flight difficulty coefficient, a path efficiency sub-model is constructed.

[0015] Based on the overlap evaluation sub-model and the path efficiency sub-model, a bi-objective optimization model for the test image sequence is established.

[0016] In some technical solutions of this application, the flight path parameters of the aerial photography equipment are obtained by solving the bi-objective optimization model, including:

[0017] Based on the limitations on image overlap and flight safety requirements, candidate paths are generated;

[0018] Based on the constraints on the image overlap and the requirements for path efficiency, a quality function for pheromone updates is generated.

[0019] The flight path parameters of the aerial photography equipment are determined based on the ant colony algorithm based on the quality function and the candidate paths.

[0020] In some technical solutions of this application, the first image sequence is obtained in the following way:

[0021] The original scene data of the first accident scene is input into the feature parameter mapping model to obtain the first shooting parameters of the first accident scene;

[0022] During the flight of the aerial photography equipment according to the flight path parameters, the first frame of the first image in the first image sequence is captured according to the first shooting parameters.

[0023] The first imaging parameters are optimized using the gradient descent method to obtain the optimized second imaging parameters;

[0024] During the flight of the aerial photography equipment according to the flight path parameters, the equipment captures the second frame of the first image in the first image sequence based on the second shooting parameters until the flight is completed.

[0025] In some technical solutions of this application, the above method also includes:

[0026] Calculate quality indicators during aerial photography;

[0027] If the quality indicators meet the preset quality requirements, the flight path parameters and / or shooting parameters are adjusted.

[0028] In some technical solutions of this application, the above method also includes:

[0029] Based on the accident characteristics at the first accident scene, determine the accident type of the first accident;

[0030] If a second accident of the same type as the first accident exists, the flight parameter set of the first accident is used to collect a second image sequence of the second accident;

[0031] The first accident scene is shown, which is reconstructed and simulated based on the second image sequence.

[0032] Secondly, embodiments of this application provide an apparatus for accident reconstruction, the apparatus comprising:

[0033] The testing module is used to determine the flight path parameters of the aerial photography equipment based on the test image sequence of the test accident acquired by the aerial photography equipment in a test scenario.

[0034] The acquisition module is used to acquire a first image sequence captured by the aerial photography device during flight according to the flight path parameters in a real-world scenario; wherein the first image sequence includes multiple first images, and each subsequent first image is captured after optimizing the shooting parameters of the previous first image;

[0035] The display module is used to display the first accident scene obtained by reconstructing and simulating based on the first image sequence.

[0036] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described accident recovery method.

[0037] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the above-described accident recovery method.

[0038] The technical solutions provided by the embodiments of this application may include the following beneficial effects:

[0039] The method of this application includes determining the flight path parameters of the aerial photography equipment based on the test image sequence of the test accident acquired by the aerial photography equipment in a test scenario; acquiring a first image sequence captured by the aerial photography equipment during the flight of the aerial photography equipment according to the flight path parameters in an actual scenario; wherein the first image sequence includes multiple first images, and each subsequent first image is obtained by optimizing the shooting parameters of the previous first image; and displaying the first accident scene obtained by reconstructing and simulating based on the first image sequence.

[0040] This application utilizes the efficient data acquisition capabilities of UAVs and the precise analysis of computer vision technology to efficiently complete the three-dimensional reconstruction and dynamic analysis of accident scenes, providing a scientific basis for accident liability determination.

[0041] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 A flowchart illustrating an accident reconstruction method provided in an embodiment of this application is shown;

[0044] Figure 2 This paper shows a schematic diagram of the overall system structure for accident reconstruction provided in an embodiment of this application;

[0045] Figure 3 A schematic diagram of an accident reconstruction apparatus provided in an embodiment of this application is shown;

[0046] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0048] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0049] It should be noted that the term "comprising" will be used in the embodiments of this application to indicate the presence of the features declared thereafter, but does not exclude the addition of other features.

[0050] The primary task of accident reconstruction is to use the driving traces left on the road at the accident scene, the collision marks on the vehicle, and the vehicle's stopping position to deduce the vehicle's speed, acceleration, heading angle, and trajectory after the collision, in order to infer the cause of the traffic accident, determine the responsibility for the accident, and take appropriate action.

[0051] Traffic accident scene investigation and analysis are crucial aspects of traffic management. Traditional accident scene investigations primarily rely on manual measurement and photographic evidence collection, which has the following problems:

[0052] Limitations of manual measurement: Measurement accuracy is limited by manual operation; the measurement process is time-consuming; it is difficult to obtain complete spatial information; and key details are easily missed.

[0053] Deficiencies in accident reconstruction: lack of accurate three-dimensional spatial information; inability to conduct precise dynamic analysis; insufficiently intuitive reconstruction of the accident process; and unconvincing analysis results.

[0054] Data storage and sharing issues: On-site data is difficult to preserve completely; post-analysis methods are limited; multi-department collaboration is inefficient; and archived materials are not systematic enough.

[0055] With technological advancements, drone aerial photography, 3D reconstruction, and vehicle dynamics simulation technologies have matured, offering possibilities for solving the aforementioned problems. However, the market currently lacks a complete solution that organically combines these technologies.

[0056] Based on this, the present application provides a method, apparatus, electronic device, and storage medium for accident reconstruction, which are described below through embodiments.

[0057] Figure 1 The diagram illustrates a flowchart of an accident reconstruction method provided in an embodiment of this application, wherein the method includes steps S101-S103; specifically:

[0058] S101. In the test scenario, the flight path parameters of the aerial photography equipment are determined based on the test image sequence of the test accident acquired by the aerial photography equipment.

[0059] S102. In a real-world scenario, during the flight of the aerial photography equipment according to the flight path parameters, the first image sequence captured by the aerial photography equipment is obtained; wherein, the first image sequence includes multiple first images, and each subsequent first image is obtained by optimizing the shooting parameters of the previous first image.

[0060] S103. Display the first accident scene obtained by reconstructing and simulating based on the first image sequence.

[0061] This application utilizes the efficient data acquisition capabilities of UAVs and the precise analysis of computer vision technology to efficiently complete the three-dimensional reconstruction and dynamic analysis of accident scenes, providing a scientific basis for accident liability determination.

[0062] The following describes some embodiments of this application in detail. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0063] This application provides an accident reconstruction scheme, the system framework of which is as follows: Figure 2As shown, the system includes a data acquisition layer, a data processing layer, a simulation analysis layer, and a result display layer. Specifically, the data acquisition layer uses drones to capture images of the accident scene; the data processing layer uses modeling software to reconstruct the 3D scene; the simulation analysis layer performs dynamic simulations using dynamics and kinematics simulation software; and the result display layer uses coding software for visualization. In detail, drones are used to photograph the accident scene along a preset flight path. The captured images are imported into modeling software, where a precise 3D model of the accident scene is generated through steps such as feature extraction, camera orientation, and dense reconstruction. Then, the 3D model is imported into dynamics and kinematics simulation software, vehicle parameters and environmental conditions are set, and vehicle collision dynamics simulation is performed. Finally, coding software is used to process the simulation data to achieve 3D scene rendering and visualization of the vehicle's trajectory.

[0064] In order to acquire more accurate image data and improve shooting efficiency during data acquisition, this application embodiment requires further limiting the flight path parameters and shooting parameters of the aerial photography equipment (drone). Specifically, to ensure shooting efficiency, this application embodiment determines the shortest flight path parameters for the drone and optimizes the shooting parameters in real time during the shooting process.

[0065] Specifically, the process of determining the shortest path flight path parameters is completed under test scenarios. That is, in this embodiment of the application, the flight path parameters of the UAV are determined by analyzing the test image sequence captured by the UAV during a test accident.

[0066] When analyzing the test image sequence, this embodiment primarily considers the image overlap and path length between adjacent frames. A bi-objective optimization model for the test image sequence is established based on the image overlap and path length between adjacent frames. Here, the bi-objectives are minimum image overlap and minimum path length. By solving the bi-objective optimization model, the flight path parameters of the aerial photography device are obtained.

[0067] When constructing the dual-objective optimization model, since it is a dual-objective optimization model, this embodiment of the application needs to first construct two sub-models for the two objectives: an overlap evaluation sub-model and a path efficiency sub-model. Specifically, based on the original overlap rate, image quality factor, and terrain factor between adjacent frames in the test image sequence, an overlap evaluation sub-model is established; based on the distance between adjacent frames in the test image sequence and the flight difficulty coefficient, a path efficiency sub-model is constructed; and based on the overlap evaluation sub-model and the path efficiency sub-model, a dual-objective optimization model for the test image sequence is established.

[0068] After obtaining the bi-objective optimization model, the flight path parameters of the UAV can be obtained by solving the bi-objective optimization model. Solving the bi-objective optimization model involves: generating candidate paths based on the constraints on image overlap and flight safety requirements; generating a pheromone update quality function based on the constraints on image overlap and path efficiency requirements; and determining the flight path parameters of the aerial photography equipment based on the ant colony algorithm using the quality function and the candidate paths.

[0069] In practice, a high-precision image sequence of the test incident site (test image sequence) is acquired. The UAV's flight path planning is based on the following optimization objectives:

[0070]

[0071] in This represents a sequence of aerial path points. Indicates the degree of overlap between adjacent aerial images; : Indicates the amount of insufficient overlap; the smaller the better. Indicates the total path length. These are weighting coefficients used to balance overlap and path length. This optimization objective has two aspects:

[0072] Insufficient to minimize overlap: Ensure sufficient image overlap.

[0073] Minimize flight path: Ensure flight efficiency.

[0074] By adjusting The value can balance these two objectives: Smaller values: Emphasis is placed on image overlap; Larger values: Focus on path length optimization. By solving this optimization problem, an aerial photography path that meets coverage requirements and has a shorter flight distance can be obtained.

[0075] First, a dual-objective optimization model based on overlap and path length is constructed. This model innovatively adopts a hierarchical structure, consisting of an overlap evaluation sub-model and a path efficiency sub-model. The two objectives are quantitatively evaluated by introducing a fuzzy comprehensive evaluation index that considers image quality and terrain factors, and an improved path cost function. Simultaneously, a scene-adaptive weight adjustment mechanism is developed to dynamically balance the weight coefficients of the two sub-models based on scene complexity, terrain undulation, and other characteristics, thus making the optimization results more consistent with practical application needs. This model significantly improves path planning efficiency while ensuring the quality of aerial photography data, reducing flight distance by 20-30% compared to traditional methods.

[0076] The overlap evaluation sub-model uses the following formula:

[0077]

[0078] in, Adjacent routes and The overall overlap evaluation value between them This represents the original overlap rate. Image quality factor For terrain factors, , , These are dynamic weighting coefficients.

[0079] The path efficiency sub-model is calculated using the following formula:

[0080]

[0081] in Distance between adjacent waypoints To account for flight difficulty factors such as altitude changes and turning angles, the final optimization objective function is:

[0082]

[0083] This model significantly improves path planning efficiency while ensuring the quality of aerial photography data, reducing flight distance by 20-30% compared to traditional methods.

[0084] Among them, the overlap target is achieved through The term L(P) represents the image overlap between adjacent flight paths, while the path length objective is represented by the overall flight distance. The weighting coefficient λ is used to balance these two objectives. Through experiments on 30 different types of test accident sites, testing different values ​​of λ from 0.1 to 0.9, it was found that when λ = 0.3, a shorter flight path can be obtained while ensuring 80% forward overlap and 70% lateral overlap.

[0085] Specifically, the improved ant colony algorithm is used to solve this optimization problem, mainly involving three aspects: designing a novel pheromone update rule, developing a safety-constrained heuristic information computation method, and implementing adaptive adjustment of algorithm parameters. The core innovation lies in designing a novel pheromone update rule:

[0086]

[0087]

[0088] in The pheromone evaporation coefficient, The overlap quality function > These are the weighting coefficients. Simultaneously, heuristic information regarding safety constraints is introduced:

[0089]

[0090]

[0091] in Distance to the obstacle For the change in height, , This is the adjustment coefficient.

[0092] The improved algorithm transforms aerial survey overlap requirements into a quality function for pheromone updates, incorporates flight safety constraints into heuristic information computation, and adaptively adjusts key parameters through an iterative process. These improvements reduce computation time by 35% and average path length by 25% while maintaining route planning quality, significantly enhancing practicality and ensuring the integrity and reliability of collected data.

[0093] The specific solution process is as follows: Optimizing aerial photography path planning needs to meet the following two main objectives: Overlap requirements: Forward overlap rate ≥ 80%, Lateral overlap rate ≥ 70%, to ensure the spatial coverage integrity between aerial images. Path length minimization: Minimize the total path length of the drone flight to save flight time and power.

[0094] The objective function to be optimized is:

[0095]

[0096] Path length, indicating flight efficiency; Image overlap optimization objective; : Weighting coefficient, used to balance the two objectives (typically adjusted to a range of 0.1 to 0.9).

[0097] Algorithm initialization: Initialize parameters, pheromone matrix The initial pheromone value between all waypoints is set to a small positive number (such as 0.1), indicating that the path has not been explored.

[0098] Heuristic information : Represents the heuristic probability of choosing a certain path, defined as:

[0099]

[0100] in: Waypoints and The horizontal distance between them; Waypoints and The horizontal distance between them; : Adjustment coefficient, used to balance the effects of horizontal distance and height difference.

[0101] Parameter settings: Number of ants Generally, 50%-100% of the total number of waypoints are used to ensure comprehensive route exploration. Pheromone evaporation rate. The typical value range is [0.1, 0.5], used to control the dissipation of pheromones over time. Pheromone Importance Factor and heuristic information importance factor The initial values ​​are set to 1 and 2, indicating that heuristic information is used preferentially. Maximum number of iterations. For example, 100 times.

[0102] Path construction:

[0103] Probabilistic path selection formula:

[0104] When constructing a path, each ant starts from the current node. Select the next node The probability of is determined by the following formula:

[0105]

[0106] : The pheromone value of the current path; Heuristic information about the path; The set of nodes that the ant has not yet visited; : Pheromone importance factor; Heuristic information importance factor.

[0107] Path constraints:

[0108] Overlap Restriction: During the route construction process, the selected route must meet the requirements for heading and lateral overlap (≥80%, ≥70%).

[0109] Flight safety: Safety constraints (such as obstacle distance and flight altitude limits) are introduced into the heuristic information to ensure that the path does not enter dangerous areas.

[0110] Path assessment:

[0111] For the path generated by each ant Calculate the following indicators:

[0112]

[0113] in Path segment The distance.

[0114] Overlap optimization objective:

[0115]

[0116] Indicates the first The actual overlap of the images Indicates the degree of overlap between targets (e.g., 80%).

[0117] Overall fitness score:

[0118]

[0119] Pheromones Update:

[0120] Local pheromone update:

[0121] When each ant constructs a path, the pheromones along the path are updated in a timely manner:

[0122]

[0123] in: When the path When selected, the pheromone intensity increases.

[0124] Global pheromone update:

[0125] After all ants have completed their path construction, select the globally optimal path. Enhance and update it:

[0126]

[0127] Pheromones are enhanced by factors that are usually positively correlated with path fitness values.

[0128] Parameter adaptive adjustment:

[0129] To improve the stability and adaptability of the algorithm, the following parameters are dynamically adjusted during the iteration process:

[0130] Ant count: As the number of iterations increases, the number of ants is gradually reduced to accelerate the convergence speed.

[0131] Pheromones volatilization rate: Use a lower value (e.g., 0.1) in the early stages to enhance exploration capabilities, and gradually increase it in the later stages to prevent over-searching.

[0132] Weight coefficient λ: dynamically adjusted according to changes in path length and overlap to adapt to different scenario requirements.

[0133] Convergence condition: The algorithm stops when one of the following conditions is met:

[0134] Reaching the maximum number of iterations The fitness value of the globally optimal path in the most recent iterations. Convergence (change is less than the threshold).

[0135] The specific flight path parameters are set as follows: flight altitude: 30-50 meters; heading overlap: 80%; lateral overlap: 70%; ground resolution: 2 cm / pixel.

[0136] After obtaining the flight path parameters, the drone is controlled to fly according to these parameters at the actual accident site. During the flight, images are taken of the accident site, resulting in a first image sequence of the accident site.

[0137] For the first frame of the first image sequence, the drone needs to capture the image according to preset first shooting parameters. These first shooting parameters are determined by the original scene data of the first accident site: the original scene data of the first accident site is input into a feature parameter mapping model to obtain the first shooting parameters for the first accident site. This original scene data includes elevation data, image data (captured using other methods), etc., and then performs terrain feature extraction (calculating elevation standard deviation, slope change rate, and terrain undulation index), obstacle distribution analysis (spatial clustering identification and density calculation), and illumination condition assessment (brightness analysis and environmental factor assessment). The processed data outputs a standardized scene feature vector, which contains quantitative indicators such as terrain complexity index, obstacle distribution characteristics, and illumination condition score. The scene features are then converted into the first shooting parameters through the feature-parameter mapping model.

[0138] After capturing the first frame of the first image, the shooting parameters corresponding to the first frame of the first image are optimized to obtain the shooting parameters for the second frame of the first image. Then, the shooting parameters for the second frame of the first image are optimized to obtain the shooting parameters for the third frame of the first image, and so on. By controlling the drone to shoot based on the shooting parameters, the first image sequence can be obtained.

[0139] For optimizing the shooting parameters, the gradient descent method is used, and safety constraints and quality requirements must be considered. Specifically, in the path optimization for accident reconstruction, the objectives are: to meet image overlap requirements: forward overlap ≥ 80%, lateral overlap ≥ 70%. Shortest path length: to minimize the flight path length. Other requirements include: flight safety (avoiding obstacles, maintaining a reasonable flight altitude).

[0140] These objectives can be represented by the following optimization function:

[0141]

[0142] in: : Path length, the total distance of the path segments. Image overlap optimization objective. : Penalty term for constraints, used to include parts that violate the constraints in the objective function. λ: Weight coefficient, used to balance path length and overlap.

[0143] Gradient descent optimization steps:

[0144] (1) Construct the objective function:

[0145] By incorporating constraints into the optimization objective, the following objective function is constructed:

[0146]

[0147] in: Waypoints and Flight distance. The first in the current path The actual overlap of the images. Target overlap (e.g., 80% on the heading, 70% on the lateral). : Used to indicate penalties for constraint violations.

[0148] (2) Transform the constraints into penalty terms:

[0149] To ensure the path meets constraints, these constraints can be embedded into the objective function as penalty functions. For example, image overlap constraints: If the heading or lateral overlap of the images does not meet the requirements (e.g., 80%, 70%), a penalty function is introduced.

[0150]

[0151] When the actual overlap Achieve the goal At that time, the penalty is 0;

[0152] If the target value is not reached, the penalty value increases as the gap widens.

[0153] Flight altitude constraints: If the altitude of the aerial photography flight If the distance exceeds the safe range (e.g., below 100 meters or above 500 meters), a penalty function is introduced:

[0154]

[0155] Minimum safe altitude. Maximum safe altitude.

[0156] Obstacle distance constraint: If the flight path is close to an obstacle, the penalty function can be designed as the reciprocal of the obstacle distance:

[0157]

[0158] Current waypoint Distance to the nearest obstacle.

[0159] (3) Gradient calculation:

[0160] Based on the optimization objective function Calculate the gradient of each variable (e.g., waypoint coordinates, flight altitude, etc.). Assume the variable is... (Representing the location of waypoints, flight altitude, etc.), the gradient is:

[0161]

[0162] For example:

[0163] Path length part:

[0164]

[0165] Overlap section:

[0166]

[0167] Penalty section:

[0168] .

[0169] (4) Gradient update:

[0170] In each iteration, the waypoint position or flight parameters are updated using gradient descent:

[0171]

[0172] Learning rate controls the update step size.

[0173] (5) Project to feasible region

[0174] To ensure that the updated variables satisfy the constraints, a projection method can be used. For example, for flight altitude constraints. ≤ ≤ After the update, Project back to feasible range:

[0175]

[0176] Convergence criterion: Optimization stops when the change in the objective function is less than the threshold or when the maximum number of iterations is reached.

[0177] After obtaining the first image sequence, a 3D scene reconstruction is performed.

[0178] After aerial data acquisition, 3D scene reconstruction is performed using modeling software. The reconstruction process is based on Structure from Motion (SfM) technology, which focuses on solving the following optimization problems:

[0179]

[0180] in, Indicates the first Projection matrix of each camera, Indicates the first The coordinates of a three-dimensional point Represents the observed coordinates of the two-dimensional image. Indicates visibility indicator; The squared pixel distance between the projected 3D point and the actual observation point is calculated. This optimization process is achieved through bundle adjustment to obtain accurate camera pose and scene 3D structure.

[0181] The reconstruction process includes the following main steps: feature extraction and matching, camera pose estimation, sparse point cloud reconstruction, dense point cloud generation, triangular mesh construction, and texture mapping.

[0182] The accuracy of scene reconstruction is evaluated using reprojection error:

[0183]

[0184] in, For actual observation points, Let n be the reprojection point, and n be the number of points participating in the evaluation. Used to calculate the average error.

[0185] Vehicle dynamics simulation: After completing the 3D scene reconstruction, vehicle dynamics simulation is performed using dynamics and kinematics simulation software. The vehicle motion model is based on multibody dynamics theory, and its equations of motion can be expressed as:

[0186]

[0187] in, For the quality matrix, Here is the damping matrix. Here is the stiffness matrix. For generalized coordinates, The external force is used. The collision process adopts an elastic collision model, and the collision force can be expressed as:

[0188]

[0189] in, For contact stiffness, The damping coefficient is... This represents the amount of deformation.

[0190] The main parameter settings during the simulation process include:

[0191] Basic vehicle parameters:

[0192] Mass distribution and moment of inertia, center of gravity position, wheelbase and wheelbase;

[0193] Suspension characteristics: Spring stiffness: Damping coefficient: Anti-roll bar characteristics;

[0194] The tire model uses the PAC2002 model:

[0195]

[0196] in, For characteristic coefficients, It is the sideslip angle.

[0197] Visualization analysis and results presentation:

[0198] Finally, encoding software is used for data visualization analysis and result presentation. For smoothing the vehicle trajectory, a Savitzky-Golay filter is employed.

[0199]

[0200] in, For the smoothed data points, These are the filter coefficients. The width is half the width of the window.

[0201] The central difference method is used to calculate velocity and acceleration.

[0202]

[0203]

[0204] in: It's location data. It's speed. It is acceleration. It is the sampling time interval, and the central difference method is used to improve the calculation accuracy.

[0205] Trajectory Reproduction:

[0206] def plot_trajectory(data):

[0207] fig = plt.figure(figsize=(12, 8))

[0208] ax = fig.add_subplot(111, projection='3d')

[0209] ax.plot3D(data['x'], data['y'], data['z'])

[0210] Dynamic parameter analysis:

[0211] def analyze_dynamics(data):

[0212] # Calculate kinematic parameters

[0213] velocity = np.gradient(data['position'], data['time'])

[0214] acceleration = np.gradient(velocity, data['time'])

[0215] # Calculate dynamic parameters

[0216] kinetic_energy = 0.5 * data['mass'] * np.sum(velocity**2, axis=1)

[0217] return velocity, acceleration, kinetic_energy

[0218] Through the above solution, this application realizes a complete workflow from on-site data collection to final visualization analysis.

[0219] In an optional implementation, this application embodiment continuously receives image quality data and current flight parameters during flight, calculates real-time quality indicators through multi-dimensional quality assessment, and automatically calculates parameter adjustment amounts based on the type of quality loss when the quality indicator is detected to be below a threshold. This module also processes historical adjustment data to optimize the adjustment strategy, achieving dynamic optimization of flight parameters through this closed-loop feedback mechanism.

[0220] Specifically, this involves multi-dimensional quality assessment, evaluating aerial images across multiple quality metrics:

[0221] Overlap Heading ≥80%, Lateral ≥70%. Calculate the difference:

[0222]

[0223] Sharpness: Based on blur detection. Exposure: Detects overexposure / underexposure. Flight Deviation: Assesses the deviation between the actual flight path and the target waypoint.

[0224] If the detection quality is below the threshold: If a certain indicator is below the preset threshold (such as overlap <80%, insufficient clarity, etc.), the adjustment process will begin.

[0225] Dynamic parameter adjustment: Calculate the adjustment amount based on the type of quality problem.

[0226] Insufficient overlap: Adjust the spacing between flight paths :

[0227]

[0228] Insufficient sharpness: Increase shooting interval or adjust focus.

[0229] Exposure issues: Dynamically adjust exposure compensation :

[0230]

[0231] Flight deviation: Correct waypoint coordinates :

[0232]

[0233] Real-time parameter adjustment: Update flight path, shooting time interval or camera parameters to ensure that the image quality of the next frame meets the standard.

[0234] In an optional implementation, after obtaining the flight parameter set of the first accident scene, this embodiment of the application saves it to a knowledge base for later use. The accident type of the first accident is also saved along with the flight parameter set.

[0235] The accident type is mainly based on the terrain features of the accident site (such as flat areas, mountains, densely populated urban areas), environmental conditions (such as light intensity and weather conditions), and mission requirements (such as accident scale and complexity).

[0236] When a second accident of the same type as the first accident exists, the flight parameter set of the first accident can be directly adopted. Under the same accident type, due to the similarity of features and task requirements, optimized flight parameters (such as flight altitude, image overlap, and flight speed) can be reused in similar scenarios. For example, flight altitude and overlap standards in flat areas usually do not need adjustment, while in complex terrain, altitude and density need to be uniformly increased. By storing the optimal parameter configurations for these scenario types in a knowledge base, the system can quickly call up parameter sets suitable for the current task, reducing computational burden while ensuring the efficiency and accuracy of aerial photography missions.

[0237] Figure 3 This application provides a schematic diagram of the structure of an accident reconstruction apparatus, which includes:

[0238] The testing module is used to determine the flight path parameters of the aerial photography equipment based on the test image sequence of the test accident acquired by the aerial photography equipment in a test scenario.

[0239] The acquisition module is used to acquire a first image sequence captured by the aerial photography device during flight according to the flight path parameters in a real-world scenario; wherein the first image sequence includes multiple first images, and each subsequent first image is captured after optimizing the shooting parameters of the previous first image;

[0240] The display module is used to display the first accident scene obtained by reconstructing and simulating based on the first image sequence.

[0241] In the test scenario, based on the test image sequence of the test accident acquired by the aerial photography equipment, the flight path parameters of the aerial photography equipment are determined, including:

[0242] Based on the image overlap and path length between adjacent frames in the test image sequence, a bi-objective optimization model for the test image sequence is established.

[0243] By solving the bi-objective optimization model, the flight path parameters of the aerial photography equipment are obtained.

[0244] The step of establishing a bi-objective optimization model for the test image sequence based on the image overlap and path length between adjacent frames in the test image sequence includes:

[0245] Based on the original overlap rate, image quality factor, and terrain factor between adjacent frames in the test image sequence, an overlap evaluation sub-model is established.

[0246] Based on the distance between adjacent frames in the test image sequence and the flight difficulty coefficient, a path efficiency sub-model is constructed.

[0247] Based on the overlap evaluation sub-model and the path efficiency sub-model, a bi-objective optimization model for the test image sequence is established.

[0248] The process of solving the dual-objective optimization model to obtain the flight path parameters of the aerial photography equipment includes:

[0249] Based on the limitations on image overlap and flight safety requirements, candidate paths are generated;

[0250] Based on the constraints on the image overlap and the requirements for path efficiency, a quality function for pheromone updates is generated.

[0251] The flight path parameters of the aerial photography equipment are determined based on the ant colony algorithm based on the quality function and the candidate paths.

[0252] The first image sequence is obtained in the following manner:

[0253] The original scene data of the first accident scene is input into the feature parameter mapping model to obtain the first shooting parameters of the first accident scene;

[0254] During the flight of the aerial photography equipment according to the flight path parameters, the first frame of the first image in the first image sequence is captured according to the first shooting parameters.

[0255] The first imaging parameters are optimized using the gradient descent method to obtain the optimized second imaging parameters;

[0256] During the flight of the aerial photography equipment according to the flight path parameters, the equipment captures the second frame of the first image in the first image sequence based on the second shooting parameters until the flight is completed.

[0257] The adjustment module is used to calculate quality indicators during aerial photography.

[0258] If the quality indicators meet the preset quality requirements, the flight path parameters and / or shooting parameters are adjusted.

[0259] The determination module is used to determine the accident type of the first accident based on the accident characteristics of the first accident scene;

[0260] If a second accident of the same type as the first accident exists, the flight parameter set of the first accident is used to collect a second image sequence of the second accident;

[0261] The first accident scene is shown, which is reconstructed and simulated based on the second image sequence.

[0262] like Figure 4As shown, this application provides an electronic device for executing the accident recovery method described in this application. The device includes a memory, a processor, a bus, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the accident recovery method described above.

[0263] Specifically, the aforementioned memory and processor can be general-purpose memory and processor, without any specific limitations. When the processor runs the computer program stored in the memory, it can execute the aforementioned accident recovery method.

[0264] Corresponding to the accident recovery method in this application, this application embodiment also provides a computer-readable storage medium storing a computer program, which is executed by a processor to perform the steps of the above-described accident recovery method.

[0265] Specifically, the storage medium can be a general-purpose storage medium, such as a removable disk or hard disk. When the computer program on the storage medium is run, it can execute the aforementioned accident recovery method.

[0266] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0267] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0268] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0269] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0270] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0271] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A method of accident recovery, characterized by, The method comprises: In a test scene, flight path parameters of the aerial device are determined according to a test image sequence of a test accident obtained by the aerial device; wherein the flight path parameters comprise aerial path points constituting the shortest flight path; In an actual scene, a first image sequence is obtained by the aerial device flying according to the flight path parameters; wherein the first image sequence comprises multiple first images, and a next first image is obtained by optimizing a shooting parameter of a previous first image; A first accident scene is reconstructed and simulated according to the first image sequence; The first image sequence is obtained by: Original scene data of the first accident scene is input into a feature parameter mapping model to obtain first shooting parameters of the first accident scene; The aerial device shoots the first image sequence according to the first shooting parameters during the flight according to the flight path parameters; The first shooting parameters are optimized by using a gradient descent method to obtain second shooting parameters after optimization; The aerial device shoots the first image sequence according to the second shooting parameters during the flight according to the flight path parameters; In the test scene, the flight path parameters of the aerial device are determined according to the test image sequence of the test accident obtained by the aerial device, comprising: A double-objective optimization model of the test image sequence is established according to an image overlap degree and a path length between adjacent images in the test image sequence; The flight path parameters of the aerial device are obtained by solving the double-objective optimization model.

2. The method of claim 1, wherein, The double-objective optimization model of the test image sequence is established according to an image overlap degree and a path length between adjacent images in the test image sequence, comprising: An overlap degree evaluation sub-model is established according to an original overlap rate, an image quality factor and a terrain factor between adjacent images in the test image sequence; A path efficiency sub-model is constructed according to a distance and a flight difficulty coefficient between adjacent images in the test image sequence; The double-objective optimization model of the test image sequence is established according to the overlap degree evaluation sub-model and the path efficiency sub-model.

3. The method of claim 1, wherein, The flight path parameters of the aerial device are obtained by solving the double-objective optimization model, comprising: A candidate path is generated according to a limitation requirement of the image overlap degree and a flight safety requirement; A pheromone update quality function is generated according to a limitation requirement of the image overlap degree and a requirement of path efficiency; The flight path parameters of the aerial device are determined according to an ant colony algorithm based on the quality function and the candidate path.

4. The method of claim 1, wherein, The method further comprises: A quality index in the aerial shooting process is calculated; If the quality index meets a preset quality requirement, the flight path parameters and / or shooting parameters are adjusted.

5. The method of claim 1, wherein, The method further comprises: An accident type of the first accident is determined according to an accident feature of the first accident scene; If a second accident of the same accident type as the first accident exists, a second image sequence of the second accident is collected using the flight parameter set of the first accident; A second accident scene reconstructed from the second image sequence is displayed.

6. An apparatus for accident recovery, characterized by The device comprises: A test module configured to determine a flight path parameter of the aerial device according to a test image sequence of a test accident acquired by the aerial device in a test scenario; wherein the flight path parameter comprises an aerial path point constituting a shortest flight path; An acquisition module configured to acquire a first image sequence captured by the aerial device in an actual scenario during flight of the aerial device according to the flight path parameter; wherein the first image sequence comprises a plurality of first images, and a next first image is captured after optimization of a capturing parameter of a previous first image; A display module configured to display a first accident scene reconstructed from the first image sequence; The first image sequence is obtained by: inputting original scene data of a first accident scene into a feature parameter mapping model to obtain a first capturing parameter of the first accident scene; capturing a first image in the first image sequence according to the first capturing parameter during flight of the aerial device according to the flight path parameter; optimizing the first capturing parameter by using a gradient descent method to obtain a second capturing parameter after optimization; capturing a second first image in the first image sequence according to the second capturing parameter during flight of the aerial device according to the flight path parameter until the flight is completed; The determination of the flight path parameter of the aerial device according to the test image sequence of the test accident acquired by the aerial device in the test scenario comprises: establishing a double-target optimization model of the test image sequence according to image overlap and path length between adjacent frames of images in the test image sequence; solving the double-target optimization model to obtain the flight path parameter of the aerial device.

7. An electronic device, comprising: It comprises: a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate through the bus, the machine readable instructions are executed by the processor to execute the steps of the accident restoration method in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, which is executed by the processor to execute the steps of the accident restoration method in any one of claims 1 to 6.

Images