摘要： 传统三帧差分目标检测算法容易发生边缘缺失现象,导致无法通过形态学处理来完全填补目标内部的空洞。为此,提出一 种边缘对比差分目标检测算法,以获取目标边缘。根据视频图像的连续性与运动目标速度的不确定性建立理论模型。对 传统边缘差分算法与固定参数的边缘对比差分算法进行分析,提出依据目标运动参数自适应获取当前模型参数的方法,从 而降低背景点的误判率并获得完整的目标边缘。实验结果表明,相对传统三帧差分算法,该算法能快速有效地从视频序列 中提取运动目标,且具有更好的鲁棒性。

关键词: 三帧差分算法, 边缘对比差分算法, 速度自适应, 运动目标检测, 形态学处理

Abstract: The traditional three-frame difference target detection algorithm is prone to the phenomenon of edge missing,which makes it impossible to completely fill the hole in the target by morphological processing.To solve this problem,an edge contrast difference target detection algorithm is proposed to obtain the target edge.A theoretical model is established based on the continuity of video images and the uncertainty of velocity of moving targets.The traditional edge difference algorithm and the edge contrast difference algorithm with fixed parameters are analyzed.A method of adaptive acquisition of current model parameters based on target motion parameters is proposed,which can reduce the error rate of background points and obtain a complete target edge.Experimental results show that,compared with the traditional three-frame difference algorithm,this algorithm can extract moving objects from video sequences quickly and effectively,and has better robustness.

Key words: three-frame difference algorithm, edge contrast difference algorithm, speed adaptive, moving object detection, morphological processing

中图分类号: 

• TP391