Title :
Adaptive least-squares technique for multi-band image enhancement
Author :
Tom, Victor T. ; Carlotto, Mark J.
Author_Institution :
Analytic Sciences Corporation, Reading, Massachusetts
Abstract :
A new image enhancement technique, based on an adaptive multi-band least-squares (LS) approach, has been developed to enhance LANDSAT-4 Thematic Mapper (TM) thermal infrared (IR) and Seasat SAR imagery. The apparent ground resolution for TM data has been improved from 120 to 30 meters and the noise power for SAR data has been decreased. The approach relies on the assumption that at a sufficiently small resolution, registered imagery data is locally correlated across bands or source sets. Using this local correlation property, visible and IR reference bands are used to predict thermal IR image data. In addition, TM reference bands are used to predict seasat SAR data.
Keywords :
Frequency; Image analysis; Image enhancement; Image resolution; Jacobian matrices; Layout; Satellites; Signal resolution; Spatial resolution; Speckle;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '85.
DOI :
10.1109/ICASSP.1985.1168319