jeromereddy9/South-African-Bank-Notes-Recognition

An image processing system for classifying South African banknotes (R10, R20, R50, R100, R200) invariant to rotation, scale, and side. Implements three models (SIFT-FLANN, ResNet-18, SimCLR) with ensemble voting. Includes preprocessing, segmentation, feature extraction, and a Streamlit GUI. # South African Bank Notes Recognition An image processing system for classifying South African banknotes (R10, R20, R50, R100, R200) invariant to rotation, scale, and side. Implements three models (SIFT-FLANN, ResNet-18, SimCLR) with ensemble voting. Includes preprocessing, segmentation, feature extraction, and a Streamlit GUI. ## How to Run the Application ### 1. Install Dependencies Download or clone the repository, then open a terminal or command prompt at that location. If you downloaded it as a ZIP file, unzip it first. Then navigate to the root repository folder: **If you downloaded and unzipped the repository:** ```bash cd South-African-Bank-Notes-Recognition-main ``` **If you cloned the repository using Git:** ```bash cd South-African-Bank-Notes-Recognition ``` Then install the required dependencies: ```bash pip install -r requirements.txt ``` ### 2. Navigate to the Application Directory Next, navigate into the inner folder containing the application script: **If you downloaded and unzipped the repository:** ```bash cd South-African-Bank-Notes-Recognition-main ``` **If you cloned the repository using Git:** ```bash cd South-African-Bank-Notes-Recognition ``` You can confirm you are in the correct location by checking that `South_African_Bank_Notes_Recognition.py` is present in the current directory. ### 3. Launch the Application Run the following command to start the application: ```bash streamlit run South_African_Bank_Notes_Recognition.py ``` Streamlit will display a local URL in the terminal (typically `localhost`) and may open a browser window automatically. If it doesn't, copy the URL and paste it into your browser.