An efficient artificial bee colony with boundary value segmentation for shadow detection

Abstract

In the area of computer vision shadow detection is important as it provides the image with high resolution with pixel clarity. There are several approaches have already be introduced but scope in this area is wide and open. In this paper an efficient artificial bee colony (ABC) with boundary value segmentation for shadow detection has been proposed. The proposed approach is the hybridization of association rules, Otsu approach, gradient segmentation and ABC algorithm. First the data preprocessing has been applied for adjoin and adjacent matrix. Then association rules have been applied for adjoin and associated pixels and joint correlations have been formed. The Otsu’s approach and gradient segmentation have been applied. It is beneficial in threshold value estimation in case of intra class variations also. Finally ABC algorithm has been applied for the final object detection and tracking. The results indicate that the approach is efficient in shadow detection and tracking.