AI-Powered Fish Monitoring: Supporting Indigenous-led Conservation in Deadman Watershed

Secwépemc Fisheries Commission (SFC) currently relies on a semi-automated system that requires biologists and technicians to manually review video footage to confirm fish counts. This process is time-consuming, delays reporting, and limits access to real-time information. The interns will develop an automated fish-counting tool that uses computer vision and artificial intelligence (YOLOv2 object detection, ByteTracker, and custom counting logic) to detect and count salmon in the Deadman Watershed, part of the traditional territory of the Skeetchestn Indian Band. By integrating the resulting dashboard into the existing infrastructure, the system will reduce the manual validation and free up staff resources to focus on higher-value conservation work. Beyond operational efficiency and technical product development, this project also strengthens Indigenous-led monitoring efforts and builds the foundation for long-term collaborative relationships between Indigenous communities and TRU.

Faculty Supervisor:

Sean Hellingman

Student:

Partner:

Secwépemc Fisheries Commission

Discipline:

Computer science

Sector:

Public administration

University:

Thompson Rivers University

Program: