The performance is not adequate, what else can I configure?
WARNING: Please read the disclaimer to make sure that all of the installation constraints have been fulfilled!
To ensure proper functioning of AI-Smoke the camera should be positioned so that the smoke to be detected has a dimension of at least 600 pixels (eg.20x30px, 15x40px).
AI-Smoke is particularly suitable for outdoor areas.
Smoke sensor configuration
In the drop-down menu on the left, by selecting "Algorithm Parameters" and then “Smoke sensor” you can add, remove and edit Smoke sensors within the scene framed by the camera.
Specifically, by clicking on “Add sensor” you can draw a virtual coil directly on the image through "Re-draw sensor", until a coil of desired dimensions is drawn.
After having added a sensor it is possible to configure the following parameters:
- Sensor ID: unique identification of added sensor
- Sensor name: name associated to the added sensor
Besides these two parameters it is also possible to configure the following to achieve an even better configuration:
- Confidence: A small value (< 0,5) makes the algorithm extremely sensitive, while a value that is too high (> 0,8) might lead to missed detections. It is advisable to choose a value between 0,5 and 0,75
- Inhibition: Inhibition time in seconds after an alarm is generated by the specific sensor.
- Latency (s): an event is generated only if the condition for the alarm persists for a certain amount of the latency time.
- RGB Threshold: difference in absokute value between the RGB channels. A value <15 makes the algorithm extremely sensitive while a value >100 creates the opposite situation.
- Saturation Threshold: a value <15 makes the algorithm extremely sensitive while a value >100 creates the opposite situation.
- Tau background slow (ms): entrance time in ms of an object in the background updating slowly.
- Tau background fast (ms): entrance time in ms of an object in the background updating fast.