Guide to Using the CBIR System

To use the CBIR system for finding similar car's images to your car photo, follow these simplified steps:

Step 1: Install Required Software

1. Install Python: Make sure Python is installed on your computer. You can download it from https://www.python.org/.

2. Install Necessary Libraries: Open your terminal or command prompt and run:

   pip install opencv-python numpy matplotlib

Step 2: Prepare Your Image Dataset

1. Organize Your Images: Create a directory and place all your JPG images in this directory. You can use the Standford cars dataset : https://www.kaggle.com/datasets/jessicali9530/stanford-cars-dataset

Step 3: Generate ORB Descriptors

1. Open Terminal or Command Prompt: Navigate to the directory containing the descriptorsGenerator.py script.

2. Run the Descriptors Generator: Execute the following command to generate ORB descriptors:

   python descriptorsGenerator.py --dataset path_to_image_dataset --output path_to_output_directory

   • Replace path_to_image_dataset with the path to your image directory.
   • Replace path_to_output_directory with the path where you want to save the descriptor files.

Step 4: Generate the Output Image with Matches

1. Run the CBIR Script: Use the CBIR.py script to generate an output image with matches. You will need to specify the query image and the dataset directory.

2. Command to Run:

   python CBIR.py --query path_to_query_image --dataset path_to_image_dataset

   • Replace path_to_query_image with the path to your query image.
   • Replace path_to_image_dataset with the path to your image dataset directory.

By following these steps, you can generate an output image that shows the matches for your query image using the CBIR system.