DRLSE-Image-Segmentation

Distance Regularized Level Set Evolution and Its Application to Image Segmentation

Level Set Evolution (LSE) is well-known method for contour extraction (determine the border of the object) and object segmentation. The main handicap of LSE is re-initialization step. This step has to be implemented to get rid of irregularities of extracted border of object (contour). Basically, level set has to be periodically re-initialize according to some distance based criteria. To the fact that how we can implement re-initialization step is not theoretical solved problem. In engineering practice, there could be significant amount of errors on to the results. In this paper, researcher proposed the new variation of LSE method which intrinsically maintains level set function instead of re-initialization step by the way of adding new term named distance regularized. That is why this new method’s name is Distance regularized LSE (DRLSE).