I have an Normalized Difference Vegetation Index(NDVI) and Enhanced vegetation index (EVI) layer and I want to re-scale their data ranges to value ranges from 0–to–1, in order to optimally stretch the contrast (information contained in the image). I have analysed the histogram of the distribution of grey values and need to omit the lowest 0.5% and the highest 0.5% of the data, however I am not too sure how to go about doing this. I tried doing this in excel (manually deleting 0.5% min and 0.5% max values) and then completing the step in ERDAS model builder. The formula used in model builder (NDVI- new min)/(new max-new min), however my results were unsuccessful.
Rescaling an NDVI or EVI from -1 to 1, to 0 to 1, uses the Rescale function (under Raster, Radiometric).
Clipping the top and bottom 0.5% is a percentage linear contrast stretch. To do this in ERDAS 2013, click on Panchromatic, then General Contrast. This brings up the Contrast Adjust window (seen below), and you can choose a variety of different methods of contrast adjustment.
You want to use Percentage LUT and set the % accordingly.
(I'm not sure of the specific steps in other versions unfortunately.)