Share Email Print

Proceedings Paper

Spatial Redundancy Reduction In High Spectral Resolution Images Using Parametric Modeling
Author(s): C. Mailhes; F. Castanie; P. Vermande
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

This paper deals with the redundancy reduction in multispectral images -covering about 200 spectral bands. The first part is an extension of an already published paper2 and is devoted to the scalar compression of each image pixel. Each pixel is an interferogram and an autoregressive model of order 10 gives a set of parameters that characterize the interferogram and are well-suited to quantization and transmission : the 10 first points, 10 reflection coefficients, and a model error vector. With scalar methods, a compression ratio of 4 or 6 is reached. In the second part, the compression is considered at the image level, on a column-by-column real-time processing basis. For classification applications, the first points of the normalized interferogram are shown to yield a discriminating vector. In the compression processing applied to each determined class, only one error and one first point vector can be transmitted. With regards to the reflection coefficients, a DPCM structure is proposed. The achieved compression ratio is shown to depend on the image and desired spectral matching. With a good trade-off between these, a ratio between 10 and 16 is obtained.

Paper Details

Date Published: 6 December 1989
PDF: 11 pages
Proc. SPIE 1154, Real-Time Signal Processing XII, (6 December 1989); doi: 10.1117/12.962393
Show Author Affiliations
C. Mailhes, GAPSE / ENSEEIHT (France)
F. Castanie, GAPSE / ENSEEIHT (France)
P. Vermande, CNES (France)

Published in SPIE Proceedings Vol. 1154:
Real-Time Signal Processing XII
J. P. Letellier, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?