University:University of Puerto Rico, Mayaguez (Puerto Rico)
Candidate:Morales Irizarry, Elsie V
Full Text:PDF
GTID:2448390002995869
Subject:Engineering
Abstract/Summary:
In this project, we implemented iterative algorithms for deblurring multispectral images. Deblurring was performed in a per channel basis. The algorithms were applied to LANDSAT-4 imagery of Puerto Rico. Deblurring methods resulted in sharpening of spatial features particularly in the infrared channels. Effects on classification were studied an it is shown that deblurring allowed correct classification of small spatial features lost due to sensor and atmospheric blurring.