I know there are a lot of threads here about using the RPi as a security camera when used with either a USB camera or the RPi camera module. I don't want to go that route. My thought was to use an RPi as the processor for multiple cameras (the heart of the security camera setup). I envision having on the order of 4-8 cameras on the outside of my house connected through either WiFi or PoE driven LAN. I would prefer fixed cameras due to expense, but the fancy PTZ dome cameras do have a definite cool factor (and associated price tag).
My main question is whether a RPi can handle motion processing for this many cameras. There are many threads documenting use of a single camera with an RPi, but none that I could locate covering any more than 2 (and 2 was just a suggestion, not implemented). If not, I suppose I'd either have to step up to a full Linux system or select one of the available systems that Samsung etc. provide. This post is mainly for input on whether an RPi is a viable option or if it is underpowered for such a task.
I plan on linking this system to my HA software to trigger lighting events most likely, so I figured that using a Linux based DIY solution might be easier to incorporate than a structured solution provided by a manufacturer. My home is prewired for a security system, but no brain was ever installed. If/when I install a brain, I will want to link it to the cameras as well.