Date Added: Jul 2012
This paper proposes an interactive tool to detect and remove shadows from colour images. The proposed method uses a Granular Reflex Fuzzy Min-Max Neural Network (GrRFMN) as a shadow classifier. GrRFMN is capable to process granules of data i.e. group of pixels in the form of hyperboxes. Granular data classification and clustering techniques are up-coming and are finding importance in the field of computer vision. Shadow detection and removal is an interesting and a difficult image enhancement problem. In this paper, a novel granule based approach for colour image enhancement is proposed. During the training phase, GrRFMN learns shadow and non-shadow regions through an interaction with the user.