Date of Award

1995

Document Type

Thesis

Publisher

Edith Cowan University

Degree Name

Bachelor of Engineering Honours

Faculty

Faculty of Science, Technology and Engineering

First Supervisor

Dr Stefan Lachowicz

Abstract

With digital motion video for either transmission or storage, there always exists a trade-off between the data transmission rate and the picture quality. The lower the transmission bit-rate is made, the more the quality of the image tends to degrade. With usual transform coding schemes the degradation usually occurs when low bit-rates, that is less than about 64Kbps, are used. The resultant image tends to suffer visually from a ''blocking" effect. This thesis therefore, is based on the development of a different implementation scheme, for motion video compression or encoding, so as to support both low bit-rates, around 64Kbps or below while eliminating the "blocking" effect. This scheme is designed round the compression of CIF, QCIF, and NTSC motion video frames, which are defined with three (one luminance and two chrominance) components per frame. These frame sizes are the same sizes used in the well-known lTV-TS standard called Recommendation H261. TI1is implementation scheme therefore closely follows that of the H.261 standard except, where such functions as the DCT transform and other modifications are needed. The Subband DCT transform used here, in replacement for the H.26l transform sub-section, is based on the work done by Yuk-Hee Chan and Wan-Chi Siu. The application of this scheme should provide similar bit-rates to that of the Recommendation H261. However, it should also provide images free of the 'blocking' effect inherent to all encoders that spatially split the image to blocks before transform coding.

Share

 
COinS