基于视频的运动车辆阴影检测与去除

Shadow Detection and Elimination of Moving Vehicles Based on Video Processing

对交通目标监控系统中运动车辆阴影的检测与去除的关键步骤进行研究,利用车辆阴影所固有的特征进行阴影抑制,针对车辆颜色与背景颜色相近时分割容易丢失目标车辆部分像素点的问题,提出了在提取的过程中将图像阈值分割和边缘检测相结合的算法,较好地把目标车辆完整的提取出来,便于车辆的准确定位.

This paper investigates the crucial steps of the shadow detection and elimination of moving vehicles in the traffic monitoring system. The essential characteristics of vehicle shadows are used to carry out shadow repression. In addition, to solve the problem of partial loss of target vehicle pixels in segmentation when the vehicle color is similar to the background color, a new algorithm integrating the shreshold-based segmentation and edge detection is presented, It will effectively extract full target vehicles and accurately position them.