Date of Submission

2-28-2008

Date of Award

2-28-2009

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Doctoral Thesis

Degree Name

Doctor of Philosophy

Subject Name

Mathematics

Department

Electronics and Communication Sciences Unit (ECSU-Kolkata)

Supervisor

Chandra, Bhabatosh (ECSU-Kolkata; ISI)

Abstract (Summary of the Work)

In the present era of multimedia, the requirement of image/video storage and transmission for video conferencing, image and video retrieval, video playback, etc. are increasing exponentially. As a result, the need for better compression technology is always in demand. The limited bandwidth of internet also asks for transmission of desired objects only. Progressive image transmission provides such facilities, where transmission is done in steps and the transmission of undesired image can be stopped at an early stage. Modern applications, in addition to high compression ratio, also demand for efficient encoding and decoding processes, so that computational constraint of many real-time applications is satisfied. In some applications, multimedia objects are compressed just once but are reconstructed frequently and most of the times the compressed objects are available on the local storage device. In such cases, the compression methods should be such that efficient reconstruction of the objects must be possible along with high quality. This thesis concentrates on image and video coding that leads to fast decoding. Such techniques are suitable for image search and retrieval and video playback.In this thesis, first of all, a novel grayscale image compression method (BTC-PF) based on block truncation coding (BTC) and vector quantization (VQ) is proposed. This method inherits the advantages of both BTC and VQ. The BTC-PF method has some controlling parameters through which we can control the bit-rate and quality. The BTC-PF method is also successfully employed to handle low bit-rate coding.The BTC-PF method is extended to color image compression. Here, the RGB model is transformed to O1O2O3 model through a lossless transformation. The BTC-PF method is then used to compress the O4 planes independently.The BTC-PF method is modified to achieve the progressive transmission of grayscaleand color images. In this respect, the patternbook which is used by BTC-PF is organized as full search progressive transmission tree (FSPTT). Finally, the residual image due to the progressive transmission is encoded by the context-based adaptive lossless image coding (CALIC) to achieve a lossless coding.In video coding, the BTC-PF method is employed for intra-frame and residual frame coding. The residual frames are usually obtained by reducing inter-frame redundancy (or temporal redundancy) through motion estimation, which is one of the most time consuming modules in a video compression system. Here also, we have proposed a fast motion estimation scheme to obtain the residual frames on which a modified version of BTC-PF method is applied. The performance of proposed methods including grayscale image compression, color image compression, progressive image transmission and video coding are quite good. As the decoding methods are mainly based on table look up, decoding module of these methods are faster than that of the state-of-the-art techniques. The effectiveness of the proposed schemes is established by comparing the performance with that of the existing methods.

Comments

ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28843462

Control Number

ISILib-TH326

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

http://dspace.isical.ac.in:8080/jspui/handle/10263/2146

Download

Included in

Mathematics Commons

Share

COinS