Image Data Compression using Zonal Sampling and Piecewise-Linear Transforms

Authors:
A. Dziech;F. Belgassem;H. J. Nern
Affiliations:
Communications Department, University of Mining and Metallurgy, Cracow, Poland;Faculty of Electronics Engineering, P.O. Box 38645, Beni-Walid, Libya;Group of Automatic Control, University of Wuppertal, Wuppertal, Germany
Venue:
Journal of Intelligent and Robotic Systems
Year:
2000

Citing 3
Cited 2

Fundamentals of digital image processing

Fundamentals of digital image processing
Digital image processing

Digital image processing
Transform Coding of Images

Transform Coding of Images

Methods of Image Compression by PHL Transform

Journal of Intelligent and Robotic Systems
Testing fusion of LDA and PCA algorithms for face recognition with images preprocessed with two-dimensional discrete cosine transform

Proceedings of the 15th WSEAS international conference on Computers

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, image data compression methods based on sample selection in the piecewise-linear transform domain will be presented. The image is subjected to a 2-dimensional piecewise-linear transformation and some coefficients will be selected using threshold method and a proposed zonal sampling method. In the proposed zonal sampling method all samples outside the selected zone will be discarded completely (no zeros are replaced). The inverse transformation in this case will have a dimension equal to that of the selected zone. Thus the number of computations needed for the inverse transformation is reduced. The Peak Signal-to-Noise Ratio (PSNR) is used as a measure of quality of the reconstructed images. Comparisons of the compression ability using piecewise-linear transforms and some selected orthogonal transforms such as Walsh and cosine transforms are given. The results obtained by using the proposed zonal sampling method show that the piecewise-linear transforms have a better performance than the orthogonal transforms.