Content adaptive deblocking of artifacts for highly compressed images

In this paper, a spatial technique for the reduction of blocking artifacts existing within the highly compressed reconstructed images, has been presented. Blocking artifacts is a kind of noise, present in images. If images are highly compressed with ‘BDCT’ based standards, then there is a possibility that some artificial boundary like artifacts may appear on reconstructed images. Removal or reduction of this type of artifacts is always desirable in image processing applications. Therefore, to remove all these artifacts, a post-deblocking method is required. In this work, to remove blocking artifacts a content-adaptive method for the deblocking has been proposed. The simulation results have been compared with other methods based on quality metrics like PSNR, MSSIM and SF where the proposed deblocking method has performed better than other methods.