README.md

metadata

title: Road Damage Detection API
emoji: 🛣️
colorFrom: blue
colorTo: green
sdk: docker
pinned: false
license: mit
short_description: AI-powered road damage detection from drone imagery

Road Damage Detection API

AI-powered highway infrastructure monitoring using YOLO-based object detection.

Features

• 🔍 Real-time Detection: Detect road damage from drone/aerial imagery
• 📍 GPS Integration: Extract and display GPS coordinates from EXIF data
• 📊 Multi-class Classification: Longitudinal cracks, transverse cracks, alligator cracks, potholes
• 🎯 Severity Assessment: Light, medium, heavy damage classification
• 📄 PDF Reports: Generate detailed inspection reports

API Endpoints

Endpoint	Method	Description
/health	GET	Health check
/detect	POST	Single image detection
/upload-survey	POST	Batch upload with GPS extraction
/analyze-survey	POST	Run AI analysis on uploaded images

Usage

# Health check
curl https://zok213-road-damage.hf.space/health

# Upload and detect
curl -X POST -F "file=@image.jpg" https://zok213-road-damage.hf.space/detect

Technology Stack

• Framework: FastAPI
• AI Model: YOLOv11 (Ultralytics)
• Image Processing: OpenCV, Pillow
• GPS: EXIF extraction (piexif)

Credits

Built by Zok213 | Da Nang, Vietnam