How harshi-max/roadvision works
RoadVision is a Streamlit pothole-detection app backed by a small road-defect processing package. Its implemented upload workflow reads video frames, optionally preprocesses them, invokes YOLO tracking, estimates severity, updates track and incident records, and produces an annotated video plus performance metrics. The README documents broader image, live-stream, mapping, export and reporting capabilities; some are not wired in the sampled UI. Edges reflect shown calls or README-described flows, not assumed integrations.
Main entry points: ui.py Streamlit workflow · incident_manager.py Incident management · roadvision Analytics and exports · report_generator.py Incident reports
Components 18
User interface
- Streamlit workflow
ui.py - Configuration
roadvision/config.py
Detection processing
- Media input
ui.py - YOLO detector and tracker
ui.py - Frame preprocessing
roadvision/preprocessing/preprocessor.py - Track records
roadvision/tracking/track_manager.py - Severity estimation
roadvision/severity/severity_estimator.py
Incident services
- Incident management
roadvision/incidents/incident_manager.py - Frame OCR
roadvision/ocr/ocr_service.py - Performance metrics
roadvision/performance/performance_monitor.py - Location providers
roadvision/location/location_provider.py - Authority routing
roadvision/authority/authority_router.py
Results and integrations
- Incident reports
roadvision/reporting/report_generator.py - Notifications
roadvision/notifications/notification_service.py - Annotated video
ui.py - Analytics and exports
roadvision - GPS map view
roadvision
Other components
- Road inspector
Connections 18
- Road inspector → Streamlit workflow: configures and starts
readme.md - Streamlit workflow → Configuration: loads settings
ui.py - Streamlit workflow → Media input: accepts upload
ui.py - Streamlit workflow → YOLO detector and tracker: tracks frames
ui.py - Streamlit workflow → Frame preprocessing: preprocesses frames
ui.py - Streamlit workflow → Track records: updates tracks
ui.py - Streamlit workflow → Severity estimation: estimates severity
ui.py - Streamlit workflow → Incident management: records detections
ui.py - Streamlit workflow → Frame OCR: samples frames
ui.py - Streamlit workflow → Performance metrics: records timing
ui.py - Streamlit workflow → Annotated video: writes annotated frames
ui.py - Streamlit workflow → Analytics and exports: aggregates detections
readme.md - Analytics and exports → GPS map view: plots GPS data
readme.md - Analytics and exports → Incident reports: exports reports
readme.md - Incident management → Incident reports: generates incident report
- Incident management → Authority routing: resolves responsible authority
- Incident management → Notifications: triggers notification
- Location providers → GPS map view: supplies coordinates
roadvision/location/__init__.py