Title :
An EZW compressed image technique for calculating texture signatures
Author :
Wilson, Beth ; Bayoumi, Magdy A.
Author_Institution :
Center for Adv. Comput. Studies, Univ. of Louisiana, Lafayette, LA, USA
Abstract :
Traditional techniques for texture feature extraction needlessly consume valuable system resources by decompressing and storing the image data prior to processing. This paper presents a new technique for calculating mean deviation texture signatures directly from wavelet-compressed images. The technique reduces the overhead while maintaining accuracy. It offers a reduction in memory and computational costs by reducing and simplifying the calculations performed and by eliminating the storage of the reconstructed image
Keywords :
data compression; feature extraction; image coding; image reconstruction; image texture; wavelet transforms; EZW compressed image technique; accuracy; computational costs; embedded zerotree wavelet coder; mean deviation signatures; memory reduction; reconstructed image; system resources; texture feature extraction; texture signatures; wavelet-compressed images; Costs; Decoding; Discrete wavelet transforms; Feature extraction; Image coding; Image reconstruction; Image storage; Peak to average power ratio; Software libraries; Web sites;
Conference_Titel :
Circuits and Systems, 2000. Proceedings of the 43rd IEEE Midwest Symposium on
Conference_Location :
Lansing, MI
Print_ISBN :
0-7803-6475-9
DOI :
10.1109/MWSCAS.2000.952828
