Eliminating target shadows for improved tracking and shape estimation in outdoor monocular diurnal sequences

We present an efficient method able to extract a shadow model from a scene, exploiting the HLS color components. The algorithm allows to recover target shapes in diurnal scene for improved identification. It is based on the realization of a General Bitmap Model and a more particular Strip Bitmap Model to identify shadow regions. Each pixel in the image is classified as shadow or not by a minimum distance approach to these models.