Hi,

I’m aware of the minimum object size for triggering a detection. The question I have is after a detection is triggered and AI analysis occurs, it seems that small objects are confirmed (shadows as “dogs”) that are smaller than the minimum object size, is this behaviour correct?

In my instance, a false detection is triggered, and ideally AI analysis filters it out, however the AI analysis picks up incorrectly a small object (smaller than minimum object size), often a combination of letter box and shadowing as a “dog” and returns a positive confirmation, however this object is smaller than my configured minimum object size.

Is it possible to define a minimum object size for the AI analysis for each camera too? Then this scenario wouldn’t occur.

Thanks

I think it is two seperate processes. BI5 has triggered on something, but it sends the entire image to CPAI, which then does its own analysis on it. I'm now getting spiders triggering my drive camera, and CPAI tells me there is a car on the drive.