Cast shadow segmentation using invariant color features

Shadows are integral parts of natural scenes and one of the elements contributing to naturalness of synthetic scenes. In many image analysis and interpretation applications, shadows interfere with fundamental tasks such as object extraction and description. For this reason, shadow segmentation is an important step in image analysis. In this paper, we propose a new cast shadow segmentation algorithm for both still and moving images. The proposed technique exploits spectral and geometrical properties of shadows in a scene to perform this task. The presence of a shadow is first hypothesized with an initial and simple evidence based on the fact that shadows darken the surface which they are cast upon. The validity of detected regions as shadows is further verified by making use of more complex hypotheses on color invariance and geometric properties of shadows. Finally, an information integration stage confirms or rejects the initial hypothesis for every detected region. Simulation results show that the proposed algorithm is robust and efficient in detecting shadows for a large class of scenes.