Double-feature based real time preceding vehicle detection
Cui Qin · Journal of Tianjin Normal University · 2010
A preceding vehicle detecting method based on shadows and symmetry without prior knowledge is presented.The shadows underneath vehicles are detected to generate vehicle hypotheses.An adaptive road threshold methold is used to get shadow feature,and a shadows area merge approach is used to resolve the problem of distortion of the shadows border.Furthermore,a method is presented to determine if there exist vehicles in the areas and to locate the vehicles based on the analysis of symmetry.The method is tested with image sequences taken on road,and the results show that the proposed algorithm can effectively detect preceding vehicles in real time.