https://www.bbc.co.uk/news/world-65972168.amp Почему аналогичные методы не используются в спутниковых источниках «больших данных» для поиска UAP Если BBC может справиться с этим, что мешает кому-либо еще с некоторыми ресурсов, чтобы сделать то же самое для UAP? «Работая с Synthetaic, компанией по искусственному интеллекту, которая просеивала огромные объемы данных, полученных со спутников, BBC нашла несколько изображений воздушных шаров, пересекающих Восточную Азию».
The government has a program that has been doing this. It’s called [Sentient](https://www.theverge.com/2019/7/31/20746926/sentient-national-reconnaissance-office-spy-satellites-artificial-intelligence-ai) . The issue is that IF UFOs move very fast then they will not appear on some of these satellite pictures. Getting clear images of planes is even difficult for these sats. Plus if a UFO is stationary for a long time then it could just be a balloon. Remember Fravor talking about the tiktac and how fast it accelerated.
What you are asking for is impossible. Why?
1. Resolution: the best commercial satellites publicly available are probably using 0.3m spatial resolution (Planet Scope). The problem is that they make acquisitions on demand (charging per 100 km2 tile).
2. You need to wait for 24 hours (at least) to get the high res data mentioned above. The acquisitions is taken in a certain timeframe (in fact in seconds) so you have available a snapshot, not a real time monitoring.
3. You aren’t able to catch fast moving objects on sat images, and I’m talking about cars, let alone UFO with 2-20 Mach speed. 😉
4. You need to know what you are looking for to train ML model. You can’t say «look for UFO», because it can look differently. And I’m simply skipping the data variation aspect like clouds, brightness, shades etc. You can use SAR acquisitions but they are pretty expensive and display s lot of noise as well.
Disclaimer. I work with AI solutions using sat images on daily basis. 😉
Because it’s already done
They are, it’s just that the US Airforce is run by LIARS