Subsample-based image compression for capsule endoscopy

Abstract

The key design challenge in capsule endoscopic system is to reduce the area and power consumption while maintaining acceptable video quality. In this paper, a subsample-based image compressor for such endoscopic application is presented. The algorithm is developed around some special features of endoscopic images. It consists of a differential pulse coded modulation followed by Golomb-rice coding. Based on the nature of endoscopic images, several subsampling schemes on the chrominance components are applied. This scheme is particularly suitable to work with any commercial low-power image sensors that outputs image pixels in a raster scan fashion, eliminating the need of memory buffer, as well as temporary storage (as needed in transform coding schemes). An image corner clipping algorithm is also presented. The reconstructed images have been verified by medical doctors for acceptability. The proposed algorithm has a very low complexity and is suitable for the VLSI implementation. Compared to other transform-based algorithms targeted to capsule endoscopy, the proposed raster-scan-based scheme performs very strongly with a compression ratio of 80% and a very high reconstruction PSNR (over 45 dB).