Title: The effectiveness of texture analysis for mapping forest land using the panchromatic bands of Landsat 7, SPOT, and IRS imagery

Author: Hoppus, Michael L.; Riemann, Rachel I.; Lister, Andrew J.; Finco, Mark V.

Year: 2002

Publication: In: Greer, Jerry Dean, ed. Rapid delivery of remote sensing products, Proceedings of the ninth forest service remote sensing applications conference; San Diego, CA; April 8-12, 2002: The American Society for Photogrammetry and Remote Sensing.

Abstract: The panchromatic bands of Landsat 7, SPOT, and IRS satellite imagery provide an opportunity to evaluate the effectiveness of texture analysis of satellite imagery for mapping of land use/cover, especially forest cover. A variety of texture algorithms, including standard deviation, Ryherd-Woodcock minimum variance adaptive window, low pass etc., were applied to moving 3x3 and 7x7 pixel windows. These windows provide quantitative information over local spatial regions that may be correlated with land cover. The three spatial resolutions of Landsat 7, SPOT and IRS panchromatic imagery, 15m, 10m and 5m respectively, provide for an analysis of pixel size when constructing cover maps. Comparison of technique effectiveness and Pan image type for providing forest cover maps was accomplished by qualitative assessment using aerial photo interpretation.