Binarization Segmentation Based on Fuzzy Clustering Theory

Abstract

Aiming at the problem of color image two value segmentation, present an improved two value plant image segmentation method based on fuzzy clustering theory. In the light of the problem that fuzzy clustering algorithm is difficult to segment the low SNR image, improve Membership function; Aiming at the problem that image pixel information is too large, mapping image information to the gray feature space, treat the same gray pixel as a whole, improve operation efficiency. The experimental results show that, the improved algorithm has stronger anti noise ability, higher segmentation accuracy, shorter operation time, has good prospects for engineering applications.