I am Using Geomatica to analyse suburban spectral signature in Las Vegas. Ran a texture analysis on all band images. Combined texture bands with original image bands to run an unsupervised ISODATA classification. I aggregated the classification to 3 classes (urban, non-urban, and suburban). I now want to compare the results of both quantitatively thus run zonal statistics. I presently at school only have access to QGIS. I assume that both my PIX files (texture analysis combined image bands, as well as classification image bands) are raster based.

Presently converting my PIX files to TIFF to make it usable in QGIS. Seeing as QGIS is my only present option to use, is it possible for me to compare two rasters here?

I am a NOOB with GIS and have no scripting knowledge. Opening the zonal statistics plugin I see only a vector vs raster option.

Does this mean I would have to convert one of my files?

I'm also not sure what would be correct in this case, do I treat my image file that contains the classification aggregation as the vector, and the texture image as the raster, or the other way around?

Lastly, both images contain heaps of data, Texture analysis is combined with 6 EMR bands, while the classification image has 14 bands (original 6, texture bands, plus 2 additional channels for the classification output and class aggregation results).

Prior to attempting zonal statistics, do I extract the classification data itself and compare solely that against the texture analysis bands?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.