eduzhai > Applied Sciences > Engineering >

A Multiparametric Class of Low-complexity Transforms for Image and Video Coding

  • Save

... pages left unread,continue reading

Document pages: 21 pages

Abstract: Discrete transforms play an important role in many signal processingapplications, and low-complexity alternatives for classical transforms becamepopular in recent years. Particularly, the discrete cosine transform (DCT) hasproven to be convenient for data compression, being employed in well-knownimage and video coding standards such as JPEG, H.264, and the recent highefficiency video coding (HEVC). In this paper, we introduce a new class oflow-complexity 8-point DCT approximations based on a series of works publishedby Bouguezel, Ahmed and Swamy. Also, a multiparametric fast algorithm thatencompasses both known and novel transforms is derived. We select thebest-performing DCT approximations after solving a multicriteria optimizationproblem, and submit them to a scaling method for obtaining larger sizetransforms. We assess these DCT approximations in both JPEG-like imagecompression and video coding experiments. We show that the optimal DCTapproximations present compelling results in terms of coding efficiency andimage quality metrics, and require only few addition or bit-shiftingoperations, being suitable for low-complexity and low-power systems.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×