Gastrointestinal endoscopy colour-based image processing technique for bleeding, lesion and reflux

摘要

This paper presents a colour-based and contour based segmentation methods to aid doctors in Gastrointestinal (GI) Endoscopy examination process. These methods are proposed to enhance existing GI Endoscopy image processing applications. Based on the previous existing GI image processing applications, these algorithms are developed to assist doctor in diagnosis process while identifying the type of disease in the intestine. For the image processing part, as compared to previous existing methods, colour-based detection is justified to be the easiest and accurate way for detecting the reflux, lesion and bleeding as they differ in colour.