So with AI, do we need to use zones at all? I currently use no zones and it runs the whole image through AI and it works great but Ive read here that you need zones for AI to work. Or is it you only need zones if you have a certain area to monitor?
Clarification on AI & Zones
- Thread starter Bosty
- Start date
sebastiantombs
Known around here
It depends on the purpose and view of the camera. For example if you have a camera watching your front door/yard area that also has the street in the background do you want to capture and analyze every time a car goes by? I use clones of cameras with specific zones for each clone to detect, and identify, specific targets..
And the capacity of your computer as well. Depending on number of cameras and field of view, you could really start pegging out the CPU sending pictures to DeepStack for every little movement...and depending on number of images and computer, you could very well max out DeepStack and then miss something.
It is best to define the areas you really want AI on. Wait until a snow storm and rain event at night and if the entire image is being sent to DeepStack, even the most robust of computers could max out CPU.
It is best to define the areas you really want AI on. Wait until a snow storm and rain event at night and if the entire image is being sent to DeepStack, even the most robust of computers could max out CPU.