Close-Range Photogrammetric Research Of Traffic Accident Reconstruction Based On The3D Camera

In recent years, with the rapid growth of economy, the amount of motor vehicleshas increased dramatically, which makes the accident happened frequently. The longtraffic accident scene investigation time can cause traffic jams, secondary accidents,and economic loss. So, to improve the speed and precision of traffic accident sceneinvestigation has become one of the goals of the modern traffic management. Therapid development of image processing and computer vision makes photogrammetrybecoming an effective way to realize this goal.To aim at the above goal, this paper analyzed the advantages and significance ofthe application of photogrammetry in traffic accident scene investigation, comparedseveral photogrammetry methods at present, summarized the development trend ofphotogrammetry in traffic accident scene: photogrammetry based on stereo cameraand3D scale, and pointed out the accuracy and efficiency of image registration,calibration and distance calculation algorithm should be improved, firstly. After that,a brief introduction of the basic theory involved in photogrammetry was given. Then,an introduction and analysis of image registration algorithm based on SIFT, which hashigher accuracy and robustness, was given. And an improved algorithm was proposed.What’s more, a summary and analysis of the distortion model and correctionalgorithm was given, and an improved distortion correction algorithm was proposed.Moreover, the paper introduced the traditional distance calculation algorithm, whichinclude extraction and recognition algorithm of identifies patterns and traditionalalgorithms of direct linear transformation, and proposed an improved distancecalculation algorithm based on stereo camera and3D scale. Finally, the summary andoutlook of this research was given.The experimental results show that the improved method has the higher accuracy,faster speed and stronger robustness than the traditional method.