M.: Shadow removal with blob-based morphological reconstruction for error correction

Dealing with shadows and highlights is essential in object detec-tion and tracking applications such as automated video surveil-lance systems for outdoors environments where accurate opera-tion is required even under variable weather conditions. In this paper, we present a novel scheme for effective shadows (high-lights) detection using both colour and texture cues. Since in any shadow removal algorithm, misclassification errors often occur, resulting in distorted object shapes. The core of this scheme is the use of a technique that is capable of correcting these errors. The technique is based on morphological reconstruction of the shadow-removed blobs conditioned on the blobs prior to the shadow-removal process, in which the object shapes are still well defined along the most part of their contours. The experiments on variety real-world video data demonstrate the favourable per-formance and robustness. 1