Project Overview

This project evaluates the environmental impact of Hurricane Maria using Sentinel-2 satellite data through NDVI analysis and object detection models. The key objectives were to identify areas most affected by the hurricane, develop a YOLO-based object detection model, and provide actionable recommendations for disaster recovery efforts.

Key Highlights:

• Conducted pre- and post-event NDVI analysis to quantify hurricane-affected areas.
• Labeled and prepared image datasets for object detection modeling using YOLO with 50 epochs.
• Developed visualizations and actionable recommendations for recovery efforts.
• Submitted a structured Jupyter Notebook with analysis, dataset preparation, and code.

This project demonstrates skills in geospatial analysis, machine learning for image processing, and disaster recovery planning.

Technologies Used

• Python: Data analysis and geospatial calculations
• YOLO: Object detection model building
• Google Earth Engine: Satellite data processing
• NDVI Analysis: Quantification of vegetation changes
• Pandas, NumPy: Data manipulation
• Matplotlib, Seaborn: Visualizations

Repository Structure

• Data/: Contains the final dataset and other relevant files.
  • submission.zip: Final dataset used for modeling and analysis.
• Code/: Includes Jupyter Notebook and other code files.
  • Business_Challenge_Analysis_and_Code.ipynb: Main Jupyter Notebook for analysis and modeling.
• Documents/: Contains the report, HTML version of the notebook, and presentation.
  • BUSINESS_CHALLENGE_III.pdf: Executive summary and findings.
  • submission.html: HTML version of the Jupyter Notebook.
• README.md: Project description and guidelines for repository usage.

Key Insights

• NDVI analysis revealed the extent of vegetation loss, highlighting critical regions for recovery planning.
• The YOLO model efficiently identified disaster-related features, aiding in situational awareness.
• Recommendations focused on targeted relief efforts, such as restoring critical green cover and infrastructure.

Instructions

1. Clone the repository to your local environment.
2. Open Business_Challenge_Analysis_and_Code.ipynb in Jupyter or any compatible environment.
3. Install required libraries, such as pandas, matplotlib, and YOLO dependencies.
4. Run the notebook to replicate analysis and modeling.
5. Review the results in the HTML version or PDF report.

Contact

For inquiries or collaboration opportunities, feel free to connect with me on LinkedIn.