A method for extracting the internal structure of biomedical images based on SRG and texture gradient

Abstract

The aim of this paper is to introduce a modified method by integrating the traditional isolated connected algorithm, which is based on SRG (Seeded Region Growing), with the texture gradient. To extend the method’s application, we not only obtained the boundary and the internal structure of 150 CT images of trabecular bone of mice, but also gained grasshoppers’ spore images which have no obscure texture. The experiment shows that this method can obtain the boundary and the internal structure of the desired part of the biomedical images quickly and reliably. It has practical value in biomedical image analysis.