Robust Watershed Segmentation of Moving Shadows using Wavelets

Segmentation of moving objects in a video sequenceis a primary mission of many computer vision tasks. However,shadows extracted along with the objects can result in largeerrors in object localization and recognition. We propose a novelmethod of moving shadow detection using wavelets andwatershed segmentation algorithm, which can effectivelyseparate the cast shadow of moving objects in a scene obtainedfrom a video sequence. The wavelet transform is used to de-noiseand enhance edges of foreground image, and to obtain anenhanced version of gradient image. Then, the watershedtransform is applied to the gradient image to segment differentparts of object including shadows. Finally a post-processingexertion is accommodated to mark segmented parts withchromacity close to the background reference as shadows.Experimental results on two datasets prove the efficiency androbustness of the proposed approach.

Keywords:

shadow removal , wavelet , watershed , chromacity based