In this job, I collaborated with Chayma Bouzaidi
In this project, we built a machine learning model to detect changes in multi-temporal satellite images.
It uses Principal Component Analysis (PCA) and K-means clustering techniques over difference image.
β’ OpenCV (version 4.2.1).
β’ Python (version 3.6.9).
β’ Scikit-learn ML Library.
β’ The directory images
contains multi-temporal images developed from the LANDSAT images available in the United States Geological Survey (USGS) website. You can find some of multi-temporal image pairs in images
directory.
β’ Go to scripts
directory and Run python DetectChange.py -io <FIRST_IMAGE> -it <SECOND_IMAGE> -o <OUTPUT_DIRECTORY>
to detect change in two multi-temporal satellite images.
NB: The output directory should end with '/'.
The script will generate a difference image named difference
and a ChangeMap
image.
Other images are generated depending on morphological transformations.