Object Segmentation based on Saliency Extraction and Bounding Box

Abstract

Object segmentation is desirable in many practical applications, e.g., object classification. However, due to various object appearances and shapes, confusing backgrounds, object segmentation in an effective way is still a challenging issue. In this paper, a novel algorithm of object segmentation based on saliency extraction and bounding boxes is proposed. The segmentation performance is significantly improved by introducing saliency extraction into the segmentation scheme. Firstly, bounding boxes are acquired by object detection algorithms, foreground and background model is constructed using bounding boxes. Then, saliency extraction procedure is introduced, and adaptive weights for each pixel are computed based on the saliency extraction. Finally, undirected graph which incorporates the adaptive weights for each pixel is constructed and graph cuts is implemented to obtain the segmentation results. Comprehensive and comparative experiments demonstrate that our proposed algorithm has achieved promising performance over a challenging public available dataset.