Moving Shadow Removal Algorithm Based on HSV Color Space

Abstract

In order to accurately separate a moving object from its shadow in a monitoring scene, this paper proposes a algorithm, which combines Multi-frame Average method for building background and HSV color space. First, the multi-frame average is used for setting up the background model. Second, the current frame and the background model are converted to HSV color space. In the foreground area, the values of brightness and saturation are smaller than that of the background, while the colors basically remain same. Just using these characters, the shadow is detected and eliminated. Finally, the algorithm is applied in videos with different monitoring scenes from the standard video library, such as highway, laboratory and campus, etc, to verify its effectiveness.  The experiment results show that the algorithm has better accuracy, reliability and robustness than the compared algorithm.

DOI : http://dx.doi.org/10.11591/telkomnika.v12i4.4991