నైరూప్య

The study of the determination of the optimal number of image segmentation based on the improved FGS method

Zhe Zhang, Xin Gao, Haizhou Ma

The important section of image analysis is the image segmentation, and the first study of image segmentation is the determination of segmentation number. The best segmentation number has comparatively effect on results of image segmentation and image analysis. This article was based on image segmentation and studied calculation model of its best segmentation image. This text summarized that on the basis of image segmentation technique and clustering analysis to talk about principle of GS AND FGS model and application in image segmentation. And this text put forward the method to determine the best image segmentation number based on improved FGS model. Through the contractive analysis of model principal, we can know that improved model has more obvious robustness on the description of image compactness than GS model and FGS model. Through image of Matlab software system and its achievement platform, the better superiority of improved model on determination of best image segmentation number than GS model and FGS model. In addition, this text did cases analysis on the effects of improved model about noise and smoothing system value to get the conclusion that noise has better effect on application of improved model, and smoothing system value has more obvious superiority on improved model than FGS model.