Difference-Index-Calculator-QGIS-Toolbox

A QGIS Processing toolbox script that calculates any normalized difference spectral index from one or two raster layers using the normalized Difference Index

Repositori di GitHub

Author: Defani Arman Alfitriansyah Date: June 11, 2026 Version: 1.0.0 GitHub: https://github.com/Defani/Difference-Index-Calculator-QGIS-Toolbox Compatible: QGIS 3.16+

Overview¶

A QGIS Processing toolbox script that calculates any normalized difference spectral index from one or two raster layers using the universal formula:

Common indices you can calculate:

This tool is part of the Remote Sensing Tools group in the QGIS Processing Toolbox — alongside the Spectral Value Extractor.

Key Features¶

• Flexible band selection — Band A and Band B can come from different raster files or from the same multi-band raster

• Auto-resample — if the two bands have different spatial resolutions, the coarser one is automatically resampled (bilinear) to match the finer resolution

• Auto-reproject — if the two rasters have different CRS, the secondary one is reprojected to match

• Custom index name — the name you enter becomes the QGIS layer name and the band description in the output GeoTIFF

• Output range clipped to [−1, 1] with NoData set to −9999

• Quick statistics printed to the log (min, max, mean)

• Output added to QGIS automatically after processing

Installation¶

Step 1 — Download the Script¶

Download NormalizedDifferenceIndexCalculator_script.py from this repository (click Code → Download ZIP or clone with git).

git clone https://github.com/Defani/Difference-Index-Calculator-QGIS-Toolbox.git

Step 2 — Add Script to Processing Toolbox¶

Open Processing → Toolbox, click the Python icon (⚙️) at the top of the Toolbox panel, and select Add Script to Toolbox...

Step 3 — Browse and Select the Script File¶

Navigate to where you saved NormalizedDifferenceIndexCalculator_script.py and click Open.

Step 4 — Find the Tool in the Toolbox¶

The tool will appear under:

Processing Toolbox
└── Scripts
    └── Remote Sensing Tools
        └── Normalized Difference Index Calculator

Double-click to open the dialog.

Usage¶

Step 5 — Fill in the Parameters¶

Step 6 — Understanding the Preset Dropdown¶

The Preset Index dropdown is a guidance tool — it shows a hint in the processing log telling you which band should go in Band A and which in Band B. You still assign the band numbers manually, which gives you full flexibility with any satellite sensor.

Example presets:

Step 7 — Auto-Resample Behavior¶

If Band A and Band B come from rasters with different pixel resolutions (e.g. 10 m vs 20 m), the tool will:

1. Detect which raster has the finer (smaller pixel) resolution

2. Resample the coarser raster to match the finer grid using bilinear interpolation

3. Compute the index on the harmonized grid

4. The output resolution matches the finest input resolution

This is reported in the processing log:

Pixel area A: 100.000000  |  Pixel area B: 400.000000
Raster B has coarser resolution (or different CRS). Resampling Band B → Band A grid (bilinear).

Step 8 — Run and Check the Log¶

Click Run. The log will show:

Preset hint: NIR band → Band A  |  Red band → Band B
Index name  : NDVI
Band A      : band 4 from 'Sentinel2_10m'
Band B      : band 3 from 'Sentinel2_10m'
Same resolution and CRS — no resampling needed.
Result  min : -0.312500
Result  max :  0.891200
Result mean :  0.423700
✔ Layer 'NDVI' added to QGIS.
✔ Done! 'NDVI' saved to: C:/output/NDVI.tif

Step 9 — Output Layer¶

Once finished, the output raster is automatically added to the QGIS Layers panel with the name you specified (e.g. NDVI). The output is a single-band Float32 GeoTIFF compressed with LZW.

Parameters Reference¶

Outputs¶

Technical Notes¶

• Output values are clipped to [−1, 1] — pixels where Band A + Band B = 0 are set to NoData

• NoData value in the output is −9999

• If both bands have the same raster source and resolution, no resampling is performed

• If CRS differ, the secondary raster is reprojected to the reference CRS using GDAL ReprojectImage with bilinear resampling

• Output GeoTIFF is written with LZW compression and tiling for efficient performance on large rasters

Requirements¶

• QGIS 3.16 or later

• Python 3.x (included with QGIS)

• Libraries: gdal, numpy (included with QGIS)

Related Tool¶

License¶

MIT License — free to use, modify, and distribute.