SUAV Image Mosaic Based on Rectification for Use in Traffic Accident Scene Diagramming

Abstract

After the traffic accident, it is a very important work on how to collect evidence from the scene, draw accident scene map, and clean up the area as soon as possible, without causing further casualties. In this paper, a method of drawing traffic accident scene map using small Unmanned Aerial Vehicles (SUAV) image mosaic based on rectification is proposed. Firstly, the accident scene images collected by SUAV are preprocessed and rectified through plane homograph. Then, two images are stitched into a large view image through feature point matching and image fusion which contains Harris corner detection and SIFT feature matching algorithm. Finally, taken the accident mosaic image as background, the accident scene map is drawn. Using this method, the scene information can be displayed correctly, which is convenient for the traffic accident scene investigators to draw the accident scene map, and ensure the complete panoramic information collection of the whole scene.