Improving the Multidimensional Multiscale Parser Algorithm for Intra Image Coding
Lucas, L.
;
Rodrigues, Nuno M. M.
; Pagliari, C.L.P.
; Silva, E.
;
Faria, S.M.M.
Improving the Multidimensional Multiscale Parser Algorithm for Intra Image Coding, Proc Conf. on Telecommunications - ConfTele, Castelo Branco, Portugal, Vol. , pp. 1 - 4, May, 2013.
Digital Object Identifier: 0
Download Full text PDF ( 176 KBs)
Abstract
In this paper, we propose some techniques to improve the Multidimensional Multiscale Parser (MMP) algorithm, specifically for intra image coding. MMP has been presented in literature as a pattern matching based algorithm that is able to outperform the rate-distortion results of the H.264/AVC standard.
Most of the proposed improvements for MMP are motivated by the new emerging multimedia systems based on very high resolution formats and multiview contents. These contents have imposed new bitrate requirements for image and video coding, which led to the research on more efficient coding techniques. We have investigated the performance of some of the new HEVC coding tools when combined with the alternative MMP algorithm.
The experimental results show that the MMP algorithm is significantly improved by using larger block sizes and a more sophisticated prediction framework. In relation to the transform-based standards, MMP demonstrates competitive results, being able to outperform HEVC for most images.