Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/46780
Title: Computation-ware block motion estimation for video coding
Authors: Rao Vinayak Sheeravanthe
Keywords: DRNTU::Engineering::Computer science and engineering::Data::Coding and information theory
Issue Date: 2009
Abstract: Computational complexity is one of the major concerns in the design of video encoders, especially for low power devices with very limited Mega instructions per second (MIPS). Conventional block-matching algorithms (BMAs) reduces the computational complexity of motion estimation by sophisticatedly inspecting a subset of checking points, stops only once all those checking points are examined and does not consider available computation power. This results in frame drop and reduction in video quality when total computation power is exhausted. Our primary goal in this project is to maximize the coding efficiency for a given computation power. A novel computation-aware scheme is proposed, which first dynamically determines the target amount of computation power allocated to a frame, and then allocates this to each block in a computation-distortion-optimized manner. The proposed computation allocation scheme can be incorporated into the widely used BMA, such as diamond, three step search (TSS), new three step search, four-step search, and hexagonal, to develop their corresponding computation-aware BMAs versions. The simulation result shows that the proposed algorithm is simple yet effective compared to an existing computation allotment scheme for software based motion estimation.
Description: 92 p.
URI: http://hdl.handle.net/10356/46780
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
EEE_THESES_129.pdf
  Restricted Access
10.21 MBAdobe PDFView/Open

Page view(s) 5

167
checked on Oct 29, 2020

Download(s) 5

6
checked on Oct 29, 2020

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.