3D accident reconstruction using low-cost imaging technique

摘要

This research is about the implementation of close range photogrammetry (CRP) technique to investigate the traffic accident scene. CRP technique representing and measuring the 3D objects using data stored in 2D photographs. It is non-contact measurement requiring multiple pictures capture to measure objects of interest. Currently, a police officer uses a conventional technique to collect data on traffic accident scene using tape measurement in order to reconstruct the accident scene. The development of camera technology increases the efficiency, stability of consumer grade digital cameras for photogram-metric applications. This development can help police officer to apply a CRP technique for data collection on traffic accident scene. CRP technique offers fast data acquisition, and only need one officer during data capture. This research is to investigate traffic accident scene using the imaging technique. The methodology of this research used a measuring tape and total station for data acquisition on traffic accident scene. This research also evaluates the accuracy of data based on mathematical model standard deviation and root mean square error (RMSE). Based on this research, the implementation of CRP in accident reconstruction is very effective because can achieve a cm-level accuracy in range 40 m traffic accident scene.