This repository contains implementation of the paper 'Mohammed El Amin, Arabi & Liu, Qingjie & Wang, Yunhong. (2016). Convolutional neural network features based change detection in satellite images. 100110W. 10.1117/12.2243798'.

The feat.py file accepts the two file paths corresponding to the two input patches as arguments.

For example to find Unstructured change in two image patches img1.PNG and img2.PNG, both located in the ./data/ folder by running

$ python feat.py ./data/img1.PNG ./data/img2.PNG