RetBul(Retrieval Building) is an online image retrieval platform applied on a custom dataset of flat buildings in Athens. RetBul's retrieval mechanism is implemented with feature extraction techniques using the SURF & SIFT image descriptors. A ranking list of the most to the less relevant to the query image is generated based on the number of inliers measured comparing query and test image.

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

You need to install the following software.

• OpenCV 3+
• Python 2.7+
• NumPy
• Pandas
• Pyjson-Tricks
• VyronasDb - Vyronas Database

You may notice two versions of the same image descriptor script.

• The [descriptor]_single.py - compares a pair of images.
• The [descriptor].py - compares an image to the reference dataset.

Only in Linux machine. Execute as a single script.

This project is licensed under the MIT License - see the LICENSE.md file for details