HSV空间特征和纹理特征的阴影检测算法研究  

A Shadow Detection Algorithm Research Based on HSV Space Features and Texture Features

作　　者：李媛[1] 张艳珠[1] 渠彪[2] 李小娟[1] 

机构地区：[1]沈阳理工大学信息科学与工程学院,辽宁沈阳110159 [2]厦门大学信息科学与技术学院,福建厦门361005

出　　处：《沈阳理工大学学报》2012年第4期11-15,共5页Journal of Shenyang Ligong University

摘　　要：为了提高检测和跟踪算法的准确性,提出了一种基于阴影在HSV空间下的特点和纹理特征的阴影检测与去除算法。该算法针对阴影区域覆盖的地面和未被阴影区域覆盖的地面纹理特征基本不变的特点,先检查序列图像中的运动区域,然后在运动区域内,根据阴影在HSV空间下的特点和纹理特征对阴影进行检测与去除,为运动目标的后续处理排除阴影干扰。实验结果表明,该算法能够很好地抑制噪声,有效检测、去除图片中运动目标的阴影。In order to improve the accuracy, this paper proposes a shadow detection and re- moval algorithm based on the features of HSV space and texture. Either the area is covered by the shadow or not, it always has the same texture feature. Based on this point, the algo- rithm firstly detect the moving area in series images, and then in moving area, based on the different characteristics of shadow when in HSV space and its texture feature, it is detected from the moving target and it is removed disturbance for the following processing. Experi- mental results show that the proposed algorithm can detect the noise and shadow and remove them effectively.

关 键 词：HSV颜色空间 纹理特征 阴影检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]