here is a basic idea on how i would like it to operate.
I will make the actual models on the weekend. Its basically like your board, the only difference is the daisy chain, and the there is only 1 power input. The omega will receive power from the servo board.
Do you want to monitor it or derive power from it, or both?
They are somewhat different problems - power supplies may hide moderate variations in input voltage, while just measuring voltage could be as simple as a divider made with high value resistors (possibly a few in series) and maybe a clamp as backup input protection.
I should have been more clear. The way I wrote it is confusing. I'm wanting to use a Omega2 /w AD converter to measure cells of the battery bank. The input to the AD will be done using a divider circuit.
My other problem is to power the Omega2 directly from that same battery bank as efficiently as possible. It's typically not recommended to use power from one of the 12V cells in the battery bank (unbalance), otherwise I could have used any cheap car adapter.
Indeed, yes, I scanned the question too quickly and thought it was about streaming from USB webcams that do their own MJPEG.
But the actual topic of trying to do video output from an MT7688 is indeed not sensible.
The cheap "not quite Android" chips used in the dongles, or a pi (if you can get the GPU actually turned on for the desired compression), do make much more sense for that... and not just in the dedicated computational paths, but in having HDMI output!