Sentinel: a Jetson-based edge camera that flags zone intrusions without cloud AI

A final-year student project at Dayananda Sagar University built Sentinel, a camera system running entirely on one NVIDIA Jetson Orin Nano Super with no cloud dependency. It detects people, tracks them across frames, and opens incidents when someone enters a monitored zone. To prevent slow models from triggering alerts based on stale frames after a reconnect, every observation is stamped with its frame time, boot ID, and stream epoch, and checked before it can drive a live decision. Incidents and alerts are stored in a local SQLite outbox so alerts still go out after a restart. The project's core components can be tested on a laptop without a camera, GPU, or model.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)
Log in to join the discussion and vote.
Log in