综合灰度和纹理特征的运动阴影检测方法

摘要

In order to improve the effectiveness and robustness of shadow detection of moving objects, an approach based on gray and texture features is proposed. Firstly, the motion regions are extracted by background subtraction. Afterwards, Fast Normalized Cross-Correlation (FNCC) is used to obtain the potential shadow regions from the motion regions. Finally, Cabor wavelet is applied to conduct texture analysis on the potential shadow regions to get the real items. The results show that the proposed method is more accurate and robust than other algorithms.%为了提高运动目标阴影检测的有效性和稳健性,提出了一种综合灰度和纹理特征的阴影检测方法.该方法通过背景差分法提取运动区域,利用快速归一化互相关函数对运动区域进行检测,获得潜在的阴影区域.然后,利用Gabor小波分析潜在阴影区域的纹理特征,得到最终的阴影区域.实验结果表明,该算法能够实时有效地进行阴影检测,并具有较强的稳健性.