Classification_Final_Project

Table of Contents

• Team
• Project Prospectus
• Report

Team

• Ilan Valencius → Hyperspectral Signature/Image Classification
• Jose Cuevas → Hyperspectral Signature/Image Classification
• Ian Dulin → Classification Verification/Research
• Alexander Ronan → Classification Verification/Research

Methods

• Classifying
  • Preprocessing contains code for parsing Landsat 8 data as well as loading a shapefile, cropping it to a bounding box (lat, lon), then exporting it. Preprocessing.m must be run before Classifying which loads the variables exported from preprocessing.m and extracts appropriate endmembers.
    • Note all Landsat8 files can be parsed and examined seperately in MATLAB but all bands must be merged into two rasters, one of full study area and one roi.
• Verification
  • Actual_part_2_check loads 1) classification of study area, 2) Reference of burned areas. It then plots and examines the similarity between the two (% overlap).

Other Notes

• Data must be obtained locally due to githubs file size requirements. CA shp perimeters are able to be uploaded.