Active contour segmentation of polyps in capsule endoscopic images

Pradipta Sasmal, Yuji Iwahori, M. K. Bhuyan, Kunio Kasugai

2018 · 20 citations · 7 references

Concepts

Abstract

Capsule endoscopy (CE) has been used as a reliable diagnostic measure for detecting polyps in the gastrointestinal (GI) tract. In this paper, we propose an automated framework for efficient segmentation of such polyps from endoscopic images. The segmentation is a challenging task owing to differences in-context in each frame. Principal componet pursuit (PCP) followed by a well-known active contour (AC) model is used in the proposed segmentation method. The PCP is used for specularity removal and background subtraction task in our method. A preprocessing algorithm is proposed to improve the performance of the PCP technique via removing bigger specular regions. Finally, the polyp regions are localized using the AC model in each frame. The experimental results show that the proposed method provides a good performance in localizing the polyps. The polyp regions, which are well illuminated in image produces better result using our proposed method.

References

7