How shivammaurya14/road-defect-detector works
This repository is a road-defect detection platform: an operator starts collection or analysis in the Flask dashboard, which coordinates camera input, GPS, inference, telemetry and saved results. The dashboard also exposes mapping, settings and incident/recording management; a separate CLI runs YOLO dashcam inference. Training and the supplied dataset are documented product capabilities, though their implementation details are not sampled. Connections involving unsampled subsystems are included only where the README documents them; source excerpts do not show all runtime wiring.
Main entry points: app.py Dashboard Runtime · script.js Dashboard UI · infer_dashcam.py Standalone Dashcam · logs Local Logs
Components 16
Capture and Inference
- Dashboard Runtime
deploy/app.py - YOLO Detection
deploy/app.py - U-Net Segmentation
deploy/app.py - Standalone Dashcam
deploy/infer_dashcam.py
Dashboard
- Dashboard UI
deploy/static/script.js - Live Defect Map
deploy/static/script.js - Mode and Settings
deploy/static/script.js
Telemetry and Records
- HUD and Telemetry
deploy/app.py - Incident Logs
deploy/app.py - Video and GPS Recording
deploy/app.py - Local Logs
logs
Training Assets
- Model Training
notebook - Labeled Dataset
dataset
Other components
- Operator
- Camera or Video
- GPS Source
Connections 20
- Operator → Dashboard UI: uses
README.md - Dashboard UI → Dashboard Runtime: controls and polls
deploy/static/script.js - Dashboard UI → Dashboard Runtime: sends browser GPS
deploy/static/script.js - Dashboard Runtime → Camera or Video: reads frames
deploy/app.py - Dashboard Runtime → GPS Source: reads location
README.md - Dashboard Runtime → YOLO Detection: runs detection
deploy/app.py - Dashboard Runtime → U-Net Segmentation: runs segmentation
deploy/app.py - Dashboard Runtime → HUD and Telemetry: updates
deploy/app.py - Dashboard Runtime → Incident Logs: creates and serves
deploy/app.py - Dashboard Runtime → Video and GPS Recording: writes video and GPS
deploy/app.py - Incident Logs → Local Logs: persists logs and images
README.md - Video and GPS Recording → Local Logs: persists recordings
deploy/app.py - Dashboard UI → Live Defect Map: shows location pins
README.md - Dashboard UI → Mode and Settings: configures modes
deploy/static/script.js - Operator → Standalone Dashcam: runs standalone
README.md - Standalone Dashcam → Camera or Video: reads frames
deploy/infer_dashcam.py - Standalone Dashcam → GPS Source: reads location
deploy/infer_dashcam.py - Standalone Dashcam → Local Logs: writes incident records
deploy/infer_dashcam.py - Operator → Model Training: trains model
README.md - Model Training → Labeled Dataset: uses labeled data
README.md