基于边缘特征和多帧差分法的运动目标检测算法

摘要

针对传统的边缘检测算法检测结果不连续、细节信息易丢失等缺点,提出了一种用于运动目标检测的边缘检测方法.首先在背景差分的基础上用平移法检测目标边缘,该方法能够提取到更为精确、细节特征更加丰富、闭合性更好的目标边缘;然后将边缘检测与多帧差分法相结合进行运动目标检测.该方法的优势在于,在目标检测的过程中不需要进行背景建模与更新.实验结果表明,该方法思路简单,易于实现,检测结果精准,实时性较好.%Aiming at the shortcomings of traditional edge detection algorithms,whose results are discontinuous and easy to lose detailed information,the paper proposes an edge detection method which can be used for moving target detection.First,the target edge is detected on the basis of background difference by translation,whose result is more accurate containing more detailed information and closing better;then the moving targets are detected by combining the edge detection with multiple frame difference.The advantage of this method is that it doesn't need background modeling and update in the whole process of target detection.Results show that this method is simple,accurate and easy to implement,and it could get a better time-consuming.