车辆视频检测及阴影去除

摘要

To solve the vehicle detection in intelligent transportation system, an approach to vehicle detection and shadow removal is proposed.A method of background subtraction and frame difference for detecting vehicles in video is introduced.Vehiches in moving video are detected with method background substraction with frame difference.Morphological techniques are used to process the moving vehicles detected, and shadow model is built and shadow is removed based on color and statistical information ofshodow.The experiments show that this method can segment the moving vehicles and removal shadow rapidly and accurately from the video.%针对智能交通系统的车辆检测问题,提出了一种运动车辆检测及阴影去除的方法.采用基于背景差与帧间差的方法来检测运动视频中的车辆,对背景差和帧间差检测出的车辆进行或运算得到运动目标.再对检测出的运动目标进行形态学处理,并结合色彩及阴影统计信息建立阴影模型,去除阴影.实验结果表明,该方法能够快速、准确地在动态视频中分割出运动车辆和阴影.