Change Detection in ArcGIS Desktop?

Question

I have two .las files derived from UAV photogrammetry. From there, in global mapper, I created a shapefile for each survey (one week apart), cropped to a shapefile.

Brought those two shapefiles into ArcMap 10.5, and used shapefile to TIN tool for each. So two shapefiles, two TINs.

Used TIN to RASTER tool to create two rasters.

Used Mimnus (3d analyst) tool to create new raster showing change. Changed classification of resultant raster to have 3 classes.

1. min to -2.9 (std dev of dataset)
2. -2.89 to 2.9 (std dev of dataset)
3. 2.91 to max

This gave me areas of cut and fill adjusted for standard deviation. UAV data isn't always perfect so needed to do this to create most likely areas of change. Raster is 32-bit float.

What I want to do now, is to somehow derive the volumes from the resultant shapes that I have. This way, I have volumes of cut and fill on a pile by pile basis given a standard deviation.

I could not achieve this using the cut fill tool because I could not figure out how to create the same map and adjust my classifications accordingly. Because the data is not perfect, when I use the cut fill tool, I get no "same" data...

First photo is the first method I outlined and the second photo is the cut fill method.

Answer 1

As commented by @MichaelStimson:

The difference between the rasters, or more correctly the absolute value, represents the volume. Assuming you have both 3d and spatial analyst license: Create a conditional (binary) raster with Con for net gain and also net loss then convert both to polygon shapes with Raster to Polygon, add a field to both called baseline (float) and calc to 0 (if not already 0) then use Polygon Volume http://resources.arcgis.com/en/help/main/10.2/index.html#/Polygon_Volume/00q90000003q000000/ to obtain gain/loss volume from baseline with the filtered difference raster.