Documentation for Axxon One 2.0. Documentation for other versions of Axxon One is available too.

Previous page Smoke detector  Video stream and scene requirements for the Smoke detector Next page

Description

Unlike standard smoke detection systems, the Smoke detector faces many issues with the scene and the background in the video image. We cannot guarantee that smoke will be detected in all cases. The Smoke detector is meant to increase the likelihood of detecting smoke. However, there can be both false alarms and failures to detect actual smoke in the camera FOV.

The Smoke detector is based on neural network, the basic principle of which is learnability.

Attention!

If the Smoke detector doesn't respond to actual fire in the camera FOV, please record a video and send it to AxxonSoft. We will update Axxon One to refine the detector (neural network training) based on the video data, and in the next versions of the product the detection operation in similar conditions will be optimal.

Add-onsFor the Smoke detector to work, you must install Neuro Pack add-on (see Installing DetectorPack add-ons)


  • No labels