Spatiotemporal Image Fusion Techniques

Installation of Python packages for Image Fusion qGIS Plugin.

Windows Operating System

1. Search and Open to OSGeo4W Shell

2. OSGeo4W Shell appears as below

3. Copy and Execute the following commands in the OSGeo4W Shell

   pip install geoAnalytics pyswarms numpy scikit-learn scipy geotiff
   

   GDAL Installation : https://www.bing.com/search?q=gdal+installation+for+windows&qs=n&form=QBRE&sp=-1&lq=0&pq=gdal+installation+for+windows&sc=10-29&sk=&cvid=568AECE00BFC4171B1596E41C8F19672&ghsh=0&ghacc=0&ghpl=

Mac Operating System

1. Open Python Console in QGIS Application

2. Execute the following python code in Python Console to get Python Path.

   import sys
   print(sys.executable)
   

3. Open terminal and go to the above specified path.

4. Execute the following command in the terminal

   /Applications/QGIS.app/Contents/MacOS/bin/python3 -m pip install geoAnalytics pyswarms numpy scikit-learn scipy geotiff
   

GDAL Installation : https://mits003.github.io/studio_null/2021/07/install-gdal-on-macos/

Installation of qGIS plugin

Step 1 : Download the "Image Fusion" Plugin folder from the below github link.

Link : https://github.com/udayRage/spatiotemporal_ImageFusion_qGIS_plugin

Step 2 : Open QGIS application

Step 3 : Select manage and install plugins

Step 4 : Select "Install from ZIP" and select the zip file

Step 5 : Click on "Install Plugin"

Reference Manual : https://github.com/Raashika214/QGIS_Plugins/blob/main/Manuals/Installation/InstallationFromZipFile.pdf

Usage manual

Step 1 : Download Image Fusion Dataset from the following link: https://data.mendeley.com/datasets/p44n3j8ym5

Step 2 : Load raster files into QGIS Application.

Step 2 : Click on Image Fusion Plugin Icon.

Step 3 : Image Fusion GUI appears as shown below.

Image Fusion Prediction

Step 1 : Click on Predict tab

Step 2 : Click on Get Layers Button.

Coarse Image at t0 : Low resolution image at t0 time stamp.

Coarse Image at t1 : Low resolution image at t1 time stamp.

Fine Image at t0 : High resolution image at t0 time stamp.

Fine Image at t1 : High resolution image at t1 time stamp.

Step 3 : Select the raster layer from the combo box.

Step 4 : Click on browse button, select output directory and enter output file name.

Step 5 : Select Algorithm from Combo Box.

Current Working Algorithms :

_ Standard HISTIF

_ Improved HISTIF

Step 6 : Enter parameter values or use recommend values.

Step 7 : Click on submit button to run the program.

Image Fusion Evaluation

Step 1 : Click on Evaluate tab.

Step 2 : Load predicted file into QGIS Application. Click on Get Layers Button.

Step 3 : Select ground truth and predicted raster layers from the combo box.

Step 4 : Click on browse button, select output directory and enter output file name.

Step 5 : Click on submit button to evaluate results.