Let's say it's 40 x ~14MP cameras fixed in to a housing.
What's stopping anyone from using higher resolution subunits and more of them.
Or, if you bolt ten of these things together do you get a 5000MP camera?
500MP x 10 ips is 5 billion pixels per second that they have to process, that is equivalent to processing 20 4K30 streams at once, even without taking into consideration the extra data you would need, then you actually need to store the data somewhere and do the facial recognition. How many of these do you think would be reasonable to have in an area?
> ensuring sufficient overlap to ensure that they can
> process the images together.
It would be much easier to avoid the issue of stitching altogether and simply process the images separately, then merge the resulting output data. As far as I'm aware, you're not going to find a GPU that can process a 500MP image efficiently.
I assume you used that word sarcastically?
From Wiktionary:
plucky (comparative pluckier, superlative pluckiest)
Having or showing pluck, courage or spirit
in trying circumstances.
Synonyms
brave
spunky
feisty
I don't forsee any scaling limits really, apart from $$$'s and power availability
I wish reporters would stop mixing up cctv cameras with facial recognition cameras. 200 million cctv cameras is a meaningless number in the context of facial recognition. How many of those cameras are capable of facial recognition? No wait, that’s way to “sciencey” and detailed for this clickbait article.
Combine that with the replacement rate of cameras, and you get two factors crossing each other and meaning that the vast majority of those 200M cameras can already be used for recognition, and the ones that can't today, will be soon.
[1] https://www.independent.co.uk/news/world/asia/china-ai-crime...
[2] https://www.bbc.com/news/world-asia-china-34592186
[3] https://www.nytimes.com/2019/04/24/technology/ecuador-survei...