Technology
HyphaeShield combines computer vision, IoT sensors, and cloud processing to detect and monitor mold risks in real-time.
Computer Vision Model
- Trained model under
Trained_Model/with classes: No_Mold, Mold, Risky_Contamination. - Optimized for edge inference and cloud validation.
- Thresholds configurable per environment.
Edge + Cloud
- Edge devices capture frames and send detections securely.
- Cloud services aggregate telemetry, run batch analysis, and generate alerts.
S3-Compatible Storage
- Image/video objects stored via S3 (Storj gateway compatible).
- Presigned uploads and secure retrieval.
MongoDB Atlas Metadata
- Detections, devices, and audit logs persisted in Atlas.
- Role-based access with JWT.