Local-adaptive blocks-based predictor for lossless image compression

Pixel intensity prediction is one of significant parts used in most lossless image compression algorithms. A new method based on local adaptation and neighboring blocks data analysis is introduced in this paper. Proposed method belongs to a class of low-complexity methods to be appropriate for realt...

Full description

Saved in:
Bibliographic Details
Published in2016 XV International Symposium Problems of Redundancy in Information and Control Systems (REDUNDANCY) pp. 92 - 99
Main Authors Novikov, Dmitriy, Egorov, Nickolay, Gilmutdinov, Marat
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2016
Subjects
Complexity theory
Context
Correlation
Image edge detection
Prediction algorithms
Prediction methods
Redundancy
Online AccessGet full text

Cover

More Information
Summary:Pixel intensity prediction is one of significant parts used in most lossless image compression algorithms. A new method based on local adaptation and neighboring blocks data analysis is introduced in this paper. Proposed method belongs to a class of low-complexity methods to be appropriate for realtime compression systems. Description of algorithm is presented as optimization problem solution. Its feasibility is also proven using presented autocorrelation experiment. This experiment allows to understand feasibility of local gradients intensities usage in proposed prediction procedure. Experimental results performed on various open image test sets show that proposed method outperforms state-of-the-art low complexity prediction methods used in modern lossless image compression standards and algorithms.
DOI:10.1109/RED.2016.7779337