Image Segmentation and Object Extraction Using Binary Partition Tree

This paper proposes a morphology based edge detection method and Binary partition tree for object extraction. Starting from an initial segmentation result generated by applying multi-iteration, multi-thresholding watershed algorithm and then applied Morphology based edge detection method. The proposed method solved the problem of undesirable over-segmentation results produced in the images and final edge detection result is one closed boundary per actual region in the image. Then intensity based region merging scheme can be exploited to merge the regions. The merging sequence can be efficiently recorded by BPT, to represent the final segmentation result with a small number of segmented regions. Experimental results demonstrate that the proposed approach can obtain a better segmentation performance as compared to existing approach of BPT from the perspective of object extraction.