Optimisation d'un algorithme de compression d'images en utilisant des méta-heuristiques
No Thumbnail Available
Date
2012
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Université Oum El Bouaghi
Abstract
The basic question addressed in this thesis is the development of compression methods, for easy archiving of images of any size with high compression ratio and minimum distortions. To design techniques, we conducted a review of the main techniques for image compression published in the literature. Among these techniques, those using wavelets, the discrète cosine transform (DCT) and fractals, appear to agree with this requirement and compression techniques réversible.
An approach was then developed to compress images is fractal compression.
Applied to the image Lena, Fruit and a sériés of images of hand written documents, this method allowed us to achieve compression ratios of 73-09 % and a PSNR of 57,18 dB. Note that the method based on genetic algorithms has reduced processing time by nearly two hundred times compared with fractal methods using iterated fonction systems. We improved this method by genetic algorithms as we have achieved compression ratios of 93,39 % and a PSNR of 51,31 whose images have been reproduced faithfully and memory space required for storage.
Description
Keywords
Compression : image, Image : stockage, Algorithme de compression