Cohu Edge AI Dome IP PTZ Camera with Edge Artificial Intelligence Analytics for Critical Infrastructure Perimeters and Access Control

The Cohu Edge AI Dome is the critical infrastructure PTZ surveillance camera that integrates computer vision artificial intelligence models directly into the camera processor — without sending video to a centralized analysis server — solving the main problem of centralized AI video surveillance systems: the bandwidth and latency bottleneck that prevents scaling analysis to hundreds of cameras and generates alerts with 2-5 second delays due to video transmission latency to the server.

The camera's edge AI processor executes four independent analysis models in real time on the local video stream: person/vehicle/object classification at 30 FPS with over 94% confidence at distances up to 150 meters with the 30x optical zoom of the 4 MP 1/1.8" optics; vehicle license plate recognition (ALPR) with plate reading while moving at speeds up to 80 km/h at 60 meters for automated access control; anomalous behavior detection — loitering, virtual line crossing, out-of-pattern speed — by object trajectory analysis without a reference database to train beforehand; and bidirectional person counting in defined zones for occupancy control in government facilities.

IEC 62443-4-2 Level 2 certification guarantees the camera cannot become a cyberattack vector against the facility network: firmware is cryptographically signed with boot-time validation, management communications use TLS 1.3 with organization PKI certificates, and configuration access is segregated from the video data network via independent VLANs. This certification is a requirement of critical infrastructure tender specifications from the Peruvian state, MINDEF, and MININT since 2024.