改进的HSV阴影去除算法研究

摘要

Since the shadow of the moving object in video sequences image may cause physical deformation, and affect the moving target detection and tracking, a thresholdless shadow removal algorithms based on HSV color space is proposed. Through analyzing the characteristics of the shadow and the background in HSV color space, and using the different characteristics of shadows and moving targets in the H, S, V three components, a thresholdless shadow elimination algorithm is proposed. Experimental results show that the proposed method can remove the shadow area well, meanwhile can maintain the integrity of the foreground target area.%针对视频序列图像中运动目标的阴影会造成运动目标的物理变形，影响运动目标的检测与跟踪等问题，提出了一种基于HSV色彩空间的无阈值阴影去除算法。该方法通过分析阴影与背景的HSV彩色空间中的特性，并利用阴影与运动目标在H、S、V三个分量中的不同特点，提出了一种无阈值的阴影消除算法。实验结果表明，该方法能够很好地去除阴影区域，同时又保持前景目标区域的完整性。