Hello ArcGIS Pro community. I'm working on creating a custom Python tool in ArcGIS Pro and I could use some advice and feedback.
I have 4 band imagery split into RGB and NDVI rasters and am trying to pass an “NDVI enhanced” RGB image to a deep learning model (Only accepts 3 channel input).
I have split out NDVI values from 0 to 1 in their own raster. These represent the values -2 to 2.
These are now in a semi-transparent overlay with the extremes in red and green, and the middle values transparent. I want to get this overlay on top of the base RGB raster and output the flattened image it as a single 3 chanel RGB raster for analysis.
I’m finding the deep learning easy compared to preparing the input raster!
Here's what I'm trying to accomplish with the tool:
- Take a base RGB raster (Mine is around 300GB – will test on smaller sample using Mask Environment setting)
- Overlay a single-band raster with values between 0 and 1 on top of the base raster, with defined symbology (either set by the layer’s symbology or a defined color ramp)
- Process in blocks
- Output the result as a new raster
The key point is that I want to use the symbology (color ramp and stretch settings) that I've already applied to the overlay raster in my ArcGIS Pro project, or need a method to define this using a custom ramp.
I appreciate any insights, tips, or code snippets you can provide. Thanks in advance for your help!
