Graduation Semester and Year

2011

Language

English

Document Type

Thesis

Degree Name

Master of Science in Electrical Engineering

Department

Electrical Engineering

First Advisor

Kamisetty R Rao

Abstract

For applications with low computational capabilities like handheld devices, it is necessary that the encoding complexity is minimal. But H.264, which is the most widely accepted video platform employs several powerful coding techniques that increase encoding complexity. Hence, the objective of this thesis is to implement an algorithm which reduces the encoding complexity by about 25%, but retains the quality of the existing intra prediction algorithm. H.264 offers nine modes for intra prediction of 4x4 luminance blocks, which includes DC prediction and eight directional modes (N4). For regions with less spatial detail, H.264 supports 16x16 intra coding, where in one of the four prediction modes (DC, vertical, horizontal and planar) is chosen for the prediction of the entire luminance component of the macro-block (N16). In addition, H.264 supports intra prediction for the 8x8 chrominance blocks which also use the similar four prediction modes as 16x16 luminance blocks (N8). The existing intra prediction algorithm uses Rate Distortion Optimization (RDO) to examine all possible combinations of coding modes. Therefore the number of mode combinations for each macro-block would be N8x (16xN4 + N16) = 4 x (16 x 9 + 4), which sums up to 592.Thus, to select the best mode for one macro-block in the intra prediction, the H.264/AVC encoder carries out 592 RDO calculations. As a result, the complexity of the encoder increases extremely. This thesis adopts a complexity reduction algorithm using simple directional masks and neighboring modes where in, the number of mode combinations are reduced to 132 at the most, with negligible loss of PSNR (peak signal to noise ratio) and bit-rate increase compared with the H.264 exhaustive search.

Disciplines

Electrical and Computer Engineering | Engineering

Comments

Degree granted by The University of Texas at Arlington

Download

Share

COinS