Detection and localization of road area in traffic video sequences using motion information and fuzzy-shadowed sets

The goal of this work is to accurately detect and localize the road area in traffic video sequences. A new approach is presented in which motion information and fuzzy-shadowed sets are used. A motion history image (MHI) is calculated from the video sequence, and then combined with background model to detect the candidate road image. The last task to get the final road image, the road localization step, is accomplished by applying a fuzzy-shadowed theoretic technique with the membership function based on both chromatic characteristics and spatial characteristics. The proposed approach was tested with recorded video under various traffic conditions and the testing results were shown to be precise in detecting and localizing the road area.