Introduction
SalmonVision is a collaborative, open-source platform that leverages cutting-edge computer vision and machine learning to monitor wild salmon populations. Developed in partnership with Indigenous communities and conservation organizations, the platform provides real-time insights into salmon abundance, species composition, and migration timing.
Key Capabilities
- Real-Time Detection — Identify and count salmon from underwater cameras, weirs, and aerial drones
- Species Classification — Distinguish between species (Chinook, Sockeye, Coho, Pink, Chum) with high accuracy
- Multi-Source Data Integration — Combine video feeds, environmental sensors, and historical data
- Community-First Design — Built to serve the needs of remote and Indigenous communities
Architecture Overview
SalmonVision consists of three main components:
- Edge Processing — Lightweight models deployed on-site for initial detection
- Cloud Platform — Centralized dashboard for data aggregation, analysis, and visualization
- API Layer — RESTful API for third-party integrations and custom workflows
Who Is This Guide For?
This guide is designed for:
- Field technicians setting up cameras and monitoring stations
- Data analysts reviewing detection results and generating reports
- Developers integrating SalmonVision into existing workflows via the API
- Community managers overseeing monitoring programs
Next: Quick Start