Digital Hardware Architectures for Exact and Approximate DCT Computation Using Number Theoretic Techniques

The discrete cosine transform (DCT) is a key tool in many image and video coding applications. Therefore significant effort is been made to realize DCT computation utilizing minimum resources in terms of hardware, at acceptable accuracy. This thesis explores the use of number theoretic techniques such as algebraic integer (AI) encoding and expansion factor methods for exact and approximate DCT computation. Novel digital hardware architectures that utilize these techniques to obtain, a) improvements in accuracy b) reductions in area and power consumption, are proposed. Review of the utilized number theoretic techniques, mathematical derivations of the proposed architectures and digital hardware implementations are provided. All designs are implemented and tested on-chip using field programmable gate array (FPGA) technology and also simulated up to synthesis level for application specific integrated circuits (ASICs). Obtained accuracy and hardware metrics are presented and compared with the current state-of-the-art. Performance of AI based DCT computation algorithms in next generation video codecs are also analysed by integrating the proposed algorithms into high efficiency video coding standard (HEVC). The proposed digital architectures have potential applications in multimedia, automaticsurveillance, geospatial remote sensing, and automotive industry.

Name thesis:degree_name

Master of Science

Level thesis:degree_level

masters

Discipline thesis:degree_discipline

Electrical Engineering

Grantor dc:publisher

University of Akron

Year dc:date

2013

Author dc:creator

• Edirisuriya, Amila

Contributors dc:contributor

• Madanayake, Arjuna

• Electrical Engineering
• Computer Engineering
• DCT
• Algebraic integers
• Digital architecture
• FPGA
• ASIC

Statement dc:rights

• unrestricted
• This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws.

Language dc:language

English

Repository record dc:identifier

http://rave.ohiolink.edu/etdc/view?acc_num=akron1363233037

OAI identifier oai:identifier

oai:etd.ohiolink.edu:akron1363233037