Turning thermal imagery into useful information
ELIAN has developed a new AI model that detects people in thermal video and brings those detections into the platform’s Vision workspace.
Thermal cameras capture heat rather than visible light, so people can stand out at night, in low light or against cluttered backgrounds where a standard camera may struggle. Automatic detection helps operators keep track of activity when visibility is limited.
The example above shows multiple people detected in the scene. Operators can view detections directly on the video, review associated events and follow detection activity over time, all in the same workspace.
Before a model goes live it is checked against labelled footage from real cameras: above, the green boxes are the hand-labelled positions and the red boxes are what the model found.
Want to explore what thermal detection could do on your site? Contact us.
This update was first shared on LinkedIn.