🚀 Detection of Craters from Orbiter High Resolution using Faster R-CNN

This project aims to automate the detection of lunar surface craters using a deep learning-based object detection model, Faster R-CNN, with ResNet-50 as the backbone. Built entirely from scratch, the model is trained on high-resolution lunar images captured by the Orbiter High Resolution Camera (OHRC) and annotated in COCO format.

📌 Project Objective

• To develop an AI-powered system that can detect and localize lunar craters from satellite imagery.
• To assist in lunar terrain analysis, mission planning, and autonomous navigation by providing accurate surface information.
• To demonstrate the use of custom-built Faster R-CNN with ResNet-50, trained from scratch without pre-trained weights.

🧠 Key Features

• ✅ Custom implementation of Faster R-CNN.
• ✅ Backbone: ResNet-50 (built from scratch).
• ✅ Dataset in COCO format containing OHRC images of the lunar surface.
• ✅ Evaluation based on Precision, Recall, mAP (Mean Average Precision).
• ✅ Visualization of detection results.
• ✅ Option for real-time inference via web interface or Gradio app.

📊 Dataset

• Source: High-resolution images from Orbiter High Resolution Camera (OHRC)
• Format: COCO (.json) with images, annotations, and categories
• Classes:
  • crater (label ID: 1)

🏗️ Model Architecture

• Backbone: ResNet-50 (custom)
• Region Proposal Network (RPN): Generates object proposals
• RoI Align + Classifier Head: Refines proposals and classifies craters

⚙️ Installation & Setup

1. Clone the Repository bash git clone https://github.com/yourusername/crater-detection-fasterrcnn.git cd crater-detection-fasterrcnn

2. Create Virtual Environment bash Copy code python -m venv venv source venv/bin/activate # or venv\Scripts\activate on Windows

3. Install Dependencies bash Copy code pip install -r requirements.txt

Training the Model bash Copy code python train.py --config configs/train_config.yaml Ensure the dataset paths and hyperparameters are correctly set in the config file.

📈 Evaluation bash Copy code python evaluate.py --weights path/to/model_weights.pth Metrics such as Precision, Recall, and mAP will be displayed.

🌐** Run the Web App (Flask)** bash Copy code cd app python app.py