Segmentation on moving shadow detection and removal by symlet transform for vehicle detection

Shadow detection and Shadow removal is major issue in computer vision group. Real research in shadow detection and removal for moving objects includes verification of moving shadow points which are many times mis-classified as moving object points in a live video sequences. A large number of shadow detection and shadow removal algorithms are introduced for images but only a few researches has been done for moving objects. A novel method for shadow detection and removal is introduced using discrete wavelet transform by symlet wavelet transform. The multi-resolution property of DWT decomposes an image into four sub bands with no loss of the spatial information. Standard deviation divide by mean by symlet transform technique is proposed. The threshold point is self-executed. The proposed method is efficient and depends on symlet wavelet coefficients.