Spectral Angle Mapper (SAM)
Maximum Likelihood
Change Detection and Stastics
Normalized Difference Vegetation Index (NDVI)
Emissivity Normalization
Thermal Atmospheric Correction
Retrieve Landsat Data (USGS)
ROIS
Isodata
After retrieving the years August 2000, 2006, and 2014 Landsat data from USGS, we continued with locating ROIS (regions of interest) per dataset
through the ENVI 64 bit program. This procedure aided the ENVI program with detecting each pixel for each necessary classification process.
We ran the ROIs through Specral Angle Mapper (also known as SAM which Uses an n-Dimensional angle calculated between two spectra tomatch pixels to classes) , Maximum Likelihood (calculates probability of a certain pixel belonging to each class, and is assigned to the class in which it has the maximum likelihood of belonging to), and Isodata (pixels grouped together based on statistics rather than user-defined class. Calculatesclass means and groups remaining pixels using minimum distance techniques). As we collected data from each, we decided that
Maximum Likelhood with a threshold of 0.00 was best fit for our change detection image, as well as statistics because the unclassifed area was dismissed. We also ran the True Color Image through the Normalized Difference Vegetation Index which is used to create one band from multispectral image to show vegetation distribution. We then progressed to wanting to analyze the thermal change within Houston by first using Thermal Atmospheric Correction which is used to compensate for atmospheric effects from data acquisition. We then proceeded to Emissivity Normalization which is used to calculate temperature from thermal radiance data.