基于HSV空间和一阶梯度的阴影剪除算法

HSV Color Space and First-order Gradient Based Shadow Suppression Algorithm

提出了一种新的基于HSV颜色信息和一阶梯度信息的阴影剪除算法,该算法运用双重准则检测阴影:准则一利用了HSV颜色空间与人观察彩色的方式十分接近的特性;准则二利用了空间上的一阶梯度信息。引入一种阴影剪除算法的评估方法,并用之对论文算法的效率进行了量化分析。实验结果表明该算法能有效剪除视频场景中运动目标的阴影。

This paper proposes an efficient shadow suppression algorithm which combines HSV color information with first-order image gradient information.The algorithm uses double rules to detect shadows:Rule 1 makes use of the character of HSV color space which corresponds closely to the human perception of color;Rule 2 utilizes first-order gradient information.A quantitative method is imported to evaluate shadow suppression algorithm and further analyse the efficiency of the proposed algorithm.Extensive experiments results on indoor and outdoor image sequences demonstrate that the proposed algorithm effectively suppresses their shadows in the scene.