Home       Chase County Overview       Makesig   Maxlike       Mindist       Piped       Cluster       Isocluster       Results       References

  


Makesig

Prior to doing any supervised classification, it was necessary to create signature files using the makesig module. Makesig extracts the pixels over all bands for each class and computes the signature statistics. The maxlike, mindist, and piped modules all use the signature files to classify pixels in the scene. The signature files are created by known locations in the image 'training sites', which can be digitized in as vector polygons over the satellite image.

The first step in processing signature files is to create training site polygons. Using the false color composite and standard color composite I on-screen digitized polygons over each feature I was classifying. I saved the vector layer polygons as crops, forest, grassland, pavement, soil, and water.

Next, I used the makesig module. I ran makesig for each vector training site polygon I had created. I indicated that the files containing the training sites were vector files and entered each separate vector training site (crops, forest, grassland, etc.). I applied the vector training sites to bands 1, 2, 3, 4, 5, and 7 of the Landsat 7 imagery. Doing this I created a signature pixel file and signature statistics file for each vector training site.


  
Home       Chase County Overview       Makesig   Maxlike       Mindist       Piped       Cluster       Isocluster       Results       References