Image segmentation algorithm based on T-junctions cues
Yanyu Qian, Fengyun Cao, Lu Wang, Xuejie Yang · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2017
To improve the over-segmentation and over-merge phenomenon of single image segmentation algorithm,a novel approach of combing Graph-Based algorithm and T-junctions cues is proposed in this paper. First, a method by L0 gradient minimization is applied to the smoothing of the target image eliminate artifacts caused by noise and texture detail; Then, the initial over-segmentation result of the smoothing image using the graph-based algorithm; Finally, the final results via a region fusion strategy by t-junction cues. Experimental results on a variety of images verify the new approach’s efficiency in eliminating artifacts caused by noise,segmentation accuracy and time complexity has been significantly improved.