基于小波边缘提取的混合运动目标检测算法
A Hybrid Moving Target Detection Algorithm Based on Wavelet Edge Extraction
-
摘要: 常用的运动目标检测算法易受到噪声、阴影等因素的影响, 检测出来的运动目标边缘比较模糊, 本文因此提出一种基于小波边缘提取的运动目标检测算法, 将小波边缘检测与帧间差分法和背景差分法相结合来检测运动目标, 采用形态学滤波和连通性分析得到准确的运动目标。实验表明, 该算法可以准确的将运动目标从视频图像序列中检测出来。Abstract: The existing moving target detection algorithms are vulnerable to noise and shadows, so the edge of the moving target detection is relatively vague. Combining the inter-frame difference with background subtraction, a wavelet edge detection method is proposed to detect moving targets. The morphological filtering and connectivity analysis are applied to obtain accurate moving target. The experiment shows that the algorithm can be adopted to detect moving targets from the video image sequence accurately.