Designing of three regions de-blocking filters in BDCT coded images at low bit rate

Abstract

Compression of images and video frames at a low bit rate significantly degrades the visual quality decoded with JPEG and MPEG standards. This paper focuses on an efficient pre and post-processing technique to address and rectify the blocking artifact. This method differentiates the blocked images or frames into different regions based upon the activity function. It efficiently removes the artifacts while preserving the edge details with the help of different adaptive filters. The proposed technique is implemented on various images and videos and results are compared with multiple existing methods using performance metrics like SSIM and GBIM. The simulation results show significant improvement in subjective quality for different images and video frames. It outperforms existing de-blocking methods