Partitional k-means clustering based hybrid DCT-Vector Quantization for image compression

Author

Mahapatra, Dheeren Ku ; Jena, Uma Ranjan

Author_Institution

Dept. of Electron. & Telecommun. Eng., Veer Surendra Sai Univ. of Technol., Burla, India

fYear

2013

fDate

11-12 April 2013

Firstpage

1175

Lastpage

1179

Abstract

In this paper, we propose Hybrid DCT-VQ, a lossy image compression technique based on Vector Quantization (VQ). In the proposed technique, we use the concept of energy compaction property of DCT to reduce the computational cost of codebook and to speed up the VQ process. This method is tested on set of standard test images for different sub-block sizes. The experimental simulation results infer some interesting outcomes in terms of execution time and compression ratio as compared to that of conventional VQ process.

Keywords

discrete cosine transforms; image coding; pattern clustering; vector quantisation; VQ process; codebook; compression ratio; energy compaction property; execution time; hybrid DCT-vector quantization; lossy image compression technique; partitional K-means clustering; standard test images; Algorithm design and analysis; Clustering algorithms; Discrete cosine transforms; Image coding; Image reconstruction; Indexes; Vectors; Clustering; DCT; Vector Quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Information & Communication Technologies (ICT), 2013 IEEE Conference on

Conference_Location

JeJu Island

Print_ISBN

978-1-4673-5759-3

Type

conf

DOI

10.1109/CICT.2013.6558278

Filename

6558278