Adaptive compression of cartographic images

based on Haar's wavelets


M. Benabdellah*, M. Gharbi, F. Regragui, N. Lamouri, E. Bouyakhf

Groupe de Recherche en Informatique, Intelligence Artificielle et Reconnaissance de Formes (GRIIARF),

Faculté des Sciences Rabat-Agdal, Morocco

* Corresponding author. E-mail:

Received : 03 November 2005; revised version accepted: 11 May 2006



     In most applications of image processing, the user is interested generally only in certain regions of the image.  In these cases, it is reasonable to consider adaptive processing of different regions in the image.

In this work, we propose an adaptive image compression technique for still images using irreversible methods.  After the selection of regions of interest, the approach consists in applying Haar's wavelets compression to these regions and classical losses JPEG compression to the context of the image.  Testing this approach on cartographic images with regions of interest produced good results.  Preliminary results revealed the superiority of adaptive compression in terms of compression ratio for comparable visual image quality.  Moreover, when the context is compressed using losses JPEG, the Haar's wavelets method on regions of interest always outperforms the reversible LZ77 method in terms of compression ratio and visual image quality.


Keywords: Image processing; Adaptive compression; Regions of interest; Haar's wavelets; JPEG; LZ77.

© 2015