How keshavgaur1/apda-mitra_project works
Project Type and Purpose
• This project is a disaster relief and response solution specifically designed to support emergency management during floods and wildfires.
• It uses satellite data (including NOAA-20’s VIIRS), AI-driven image processing (semantic segmentation for flood extent analysis), and drone imagery processed with YOLOv8 for real-time victim detection.
• The system is meant to enhance situational awareness, provide early warnings, and optimize resource allocation during disaster scenarios.
Analyzing the File Structure
• Top-level Files and Folders:
• “app.py”: Likely the main application entry point, which orchestrates or ties together the various components.
• “assets/”: Contains images used in the UI or for demonstration (e.g., floods, wildfire data, victim detection visualizations).
• “Flood_mapping/”: Contains notebooks for flood detection and segmentation, which indicate the use of AI for evaluating flood impact.
• “detectFloodVictims/”: Holds notebooks and subdirectories (runs, weights, test_images) for YOLO-based detection of victims; this implies a computer vision pipeline.
• “detectWildFire/”: Contains a notebook for wildfire detection, further indicating specialized processing for satellite imagery.
• “pages/”: Contains Python scripts (e.g., 🆘save_victims.py, 🌊floods.py, 👨🚒wildfire.py) that are probably part of a front-end or API routes that drive user-facing features.
• “requirements.txt” and “packages.txt”: List dependencies that support various aspects of the system.
• The structure suggests a mixed architecture where distinct modules manage specialized processing tasks (flood segmentation, victim detection, wildfire detection) along with an interface layer.
Key Architectural Components
a. Data Sources:
• Satellite Imagery Feed: Utilizes NOAA-20 VIIRS sensors (especially for wildfire detection and flood mapping).
• Drone Imagery: For real-time, on-ground rescue operations (victim detection).
b. Data Processing Modules:
• Flood Mapping Module (Flood_mapping): Includes AI segmentation for risk assessment.
• Victim Detection Module (detectFloodVictims): Employs YOLOv8-based object detection to identify stranded victims.
• Wildfire Detection Module (detectWildFire): Processes satellite imagery to flag early-stage wildfires.
c. Application Entry and Routing:
• The main application (app.py) possibly integrates the processed data and provides API endpoints.
• “pages/” directory components might be used for rendering results to end users (web dashboard or similar).
d. Storage/Asset Management:
• “assets/”: Stores static images/visualizations that are referenced in the interface or documentation.
e. Output and Communication:
• A visual dashboard or web UI (leveraging components in the pages folder) provides real-time alerts and visualizations to disaster responders.
Specific Elements
• For each component, include technology/tool hints (e.g., YOLOv8 in Victim Detection, semantic segmentation in Flood Mapping).
• Aggregation point (app.py) connecting the processing results to user-facing dashboards.
• Flood Mapping (from Flood_mapping folder)
• Victim Detection (from detectFloodVictims)
• Wildfire Detection (from detectWildFire)
• Finally, at the right side, add the “User Interface/Dashboard (pages/)” that delivers alerts, maps, and real-time information.
• “assets/” as part of static content storage accessible by the dashboard.