Dissertation > Excellent graduate degree dissertation topics show

The Application of Deformable Block Matching Algorithm

Author: RenLiang
Tutor: HouZhengXin
School: Tianjin University
Course: Signal and Information Processing
Keywords: Video Compression Motion estimation and compensation Deformable block matching algorithm (DBMA) MPEG-2
CLC: TN919.8
Type: Master's thesis
Year: 2007
Downloads: 32
Quote: 0
Read: Download Dissertation

Abstract

This article will be a completely new motion estimation and compensation algorithm - based deformable block matching motion estimation and compensation algorithm (DBMA) into the current actual video compression coding standard MPEG-2, and the overall video compression codec adjustment and optimization of the system. This paper first introduces video compression coding standard MPEG-2 basic characteristics and the underlying knowledge, and focuses on the basic principles of traditional motion estimation and compensation algorithm; then analyzed the actual properties of the video signal from the fundamental interpretation of the traditional block-matching algorithm (BMA) are based on the model of the defects of the pan, a brief introduction for actual movement characteristics deformable block motion model based on the nodal displacements based deformable block matching motion estimation and compensation algorithms (DBMA); then DBMA into MPEG-2 video coding system, the due and DBMA increase the motion parameters, the encoder and decoder are made corresponding adjustments, so that the entire codec system can be completed independently of the source of the entire video signal encoding and decoding functions; Finally, we some optimization, such as the right the DBMA and the codec system PMVFAST_Plus calculation process optimization Two_Mode, SGE, CS, the edge filter and half-pixel accuracy search, etc., also integrated into the MPEG-2 codec, and combined to generate a new for MPEG-2 video compression codec motion estimation and compensation algorithm, referred to herein as Fast_DBMA PMVFAST_Plus TM SGE CS Egde_Filter algorithm makes the whole codec system more efficient video compression performance. The test results show that, in the case of the integer pixel search precision, under the premise of the same speed of operation and the same target bitrate Fast_DBMA peak signal-to-noise ratio (PSNR) of the reconstructed image of the PMVFAST_Plus the TM sge the CS Egde_Filter algorithm than the conventional block-based matching reconstruction of the global search algorithm (BMA) algorithm image peak signal-to-noise ratio (PSNR) on average 1.86dB, significant improvement in subjective visual effect; we were also analyzed from the simple pursuit of speed and angle of the simple pursuit of quality, also similarly satisfactory fusion results. In addition, in the case of a half-pixel search accuracy, as long as we appropriately adjusted and optimized combinations and associated parameters, Fast_DBMA PMVFAST_Plus TM Egde_Filter algorithm can still break the objective quality limit for the half-pixel accuracy global search algorithm, while the subjective visual effect is also improved ; In addition, the terms of the global search algorithm based on half-pixel precision, the algorithm for the reconstruction of Fast_DBMA FS TM Egde_Filter image PSNR PSNR of a reconstructed image about global search algorithm than the half-pixel accuracy can be higher than 0.20dB, subjective visual effect improved.

Related Dissertations

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Communicate > Image communication, multimedia communication
© 2012 www.DissertationTopic.Net Mobile