Efficient reduction of artifact effect based on power and entropy measures
Ming-Te Wu · 2015
Using the discrete cosine transform, such as JPEG or JPEG 2000, is still one of the best technologies for image and video compression standards. JPEG/JPEG2000 standards have been supplied to image compression, but this area is still open algorithms. It can provide better compression ratio while maintaining a low mean square error. A multiple scale sub-band analysis method in this study was proposed, which uses the energy and entropy metrics. The developed method can not only be used to reduce the effect of artifacts like blocking or blurred edge, compared with the JPEG 2000 and other recent developed scheme, but also to improve the PSNR performance. Simulated results are implemented to show that the proposed algorithms obtain better improvement than JPEG, JPEG 2000 and other schemes. The evaluation of this method in comparison with other approaches is implemented both subjectively and qualitatively and the simulation exemplify that the proposed algorithms could obtain the desired reconstructive performance subjectively and objectively.