Skip to Main Content
The capsule endoscopy (CE) has gradually found its applications in hospitals because of its great advantage that it can view the entire small bowel with almost no invasiveness compared to traditional endoscopes and other imaging techniques for gastrointestinal diseases. However, a major issue with this new technology is that too many images produced by CE causes a huge burden to physicians, so it is very meaningful to help the clinicians if we could partially implement computer aided diagnosis. We proposed a new scheme aimed for ulcer region discrimination in CE images in this paper. This new scheme utilizes texture feature, a very powerful clue for the physicians to diagnose, to recognize ulcer regions with neural network classifier. We advance a new idea of curvelet based local binary pattern as the textural features to discriminate ulcer regions from normal regions, which make full use of the curvelet transform and uniform local binary pattern. Experiments on our present image data sets validate that it is promising to employ the proposed texture features to recognize the ulcer regions.