Identification and Test Analysis of Scattered Objects in Traffic Accident Scene

Abstract

The scattered objects contain a lot of important information on the scene of traffic accidents. The identification and extraction of relevant characteristic values of the scattered objects on the scene of traffic accidents can provide important basis for solving the collision speed of the vehicles in the accident and provide some evidence for the handling of traffic accidents. In this paper, scattered objects on the scene of traffic accidents are taken as the research object. According to the theoretical model of camera imaging and the characteristics of scattered objects in the accident site, a camera calibration system was proposed by combining with MATALB software. In order to reduce the error caused by background environment in the calibration process, the red calibration object with known structure size was designed to obtain the aerial photography image containing the calibration object. MATLAB was used to extract the calibration object in the image and establish the relationship model between the actual value of the calibration object and the image value. The image processing accuracy was improved by gray transformation, Wiener filter denoising, Laplace sharpening and Canny algorithm edge detection. Taking circular and irregular scattered objects as examples, the simulation test is designed. The proposed calibration method and image processing method are applied to identify the features of scattered objects. The identification results are calculated and compared with the actual measurement results to verify the accuracy of the proposed method and provide a reference for on-site accident treatment.