The segmentation of glandular cavity based on K-means and mathematical morphology
Yingjun Ma, Muhammad Umair Hassan, Dongmei Niu, Liping Wang · 2017
This paper presents a new method for the segmentation of glandular cavity. Our method is based on the K-means and mathematical morphology. We have segmented the entire glandular cavity and eliminated many of the interference factors in the gastric glandular image. An important characteristic of our method is that we combined the K-means with the mathematical morphological processing, and carried out the iterative execution which resulted in a gradual refinement, and the algorithm can run in nearly a linear time. Images that are processed using our methods can be more effective in helping doctors diagnose disease.