PSSIFence is an artificial intelligence system currently in development by Fisheries and Oceans Canada that automatically detects and counts salmon species from underwater camera footage at fish fence installations. Traditionally, DFO scientists manually review hours of underwater video to count and identify fish species passing through fences placed across rivers and streams. As monitoring obligations increase, manual video review has become unsustainable.
This AI system uses advanced computer vision technology to replace manual analysis, automatically processing fence videos to determine salmon counts by species. The system aims to reduce the manual effort required while providing more accurate and timely data for stock assessment and population monitoring. The system does not involve processing personal information.
The AI system is being developed by the Government of Canada and is primarily used by GC employees in the Science Branch - Stock Assessment Lab. This system is subject to federal AI notification requirements and transparency commitments.
