GPS Coordinate Extractor (EXIF & OCR Engine)

Delivered scope

• Accepts direct Google Drive folder links or ZIP uploads with automated nested folder traversal.
• Performs server-side EXIF inspection with Tesseract OCR fallback for text-stamped coordinates.
• Generates downloadable Excel spreadsheets with coordinates, extraction source, and error summaries.

Technology

System approach

• Web Core: Flask web application with Gunicorn multi-threaded workers, live SSE progress logging, and Docker Compose orchestration.
• Ingestion Pipeline: Recursive Google Drive downloader and ZIP extraction engine supporting nested agency folders.
• Coordinate Engine: Two-tier extraction pipeline using Piexif for EXIF metadata with Tesseract OCR fallback parsing image text overlays.
• Report Generator: OpenPyXL workbook builder compiling per-photo coordinates, source method, validation status, and summary counts.

Challenges

• Recovering coordinates when camera apps or messaging platforms (e.g. WhatsApp) strip raw EXIF metadata.
• Processing massive Google Drive folders containing hundreds of high-resolution images without memory exhaustion or timeouts.
• Filtering out noise and parsing variable coordinate formats from camera timestamp text overlays.

Solutions

• Engineered a Tesseract OCR fallback pipeline using regex pattern matching for DMS and decimal coordinate overlays.
• Implemented streaming job queues with configurable TTL cleanup (6 hours) to prevent disk and memory bloat on the VPS.
• Added Google Drive link ingestion that downloads nested hierarchies automatically with real-time browser progress updates.

Key outcomes