Cell Image Decomposition Based on G-Space and PDE

摘要

Base on G-Space and partial differential equation, a novel method is proposed for cell image decomposition in this paper. The Vese-Osher (VO) model and Mumford-Shah-G(MSG) model based on G-Space and PDE are introduced for real texture image extraction. An image f can be decomposed into a cartoon part u and a noise/texture part v in both two models. The experimental results show the contrast effects in these two models for cell image decomposition. As the basis research for cell image recognition, this image decomposition approach is proved available and effective by the experimental results.