A new level set method of image segmentation is proposed in this paper, which utilizes structure tensor information. Based on analyzing the classic Region-Scalable Fitting (RSF) model in detail, a structure tensor drawing upon gray information is added to RSF model in order to overcome the weakness that RSF model cannot deal with texture segmentation. Experimental results show that the new model can obtain better segmentation effects for texture segmentation and improve the segmentation accuracy.
Published in:
Intelligent Systems and Applications (ISA), 2010 2nd International Workshop on
Date of Conference: 22-23 May 2010