Show HN: UFO Detector – Capturing Aerial Anomalies Using Computer Vision(github.com)
github.com
Show HN: UFO Detector – Capturing Aerial Anomalies Using Computer Vision
https://github.com/UFOID/UFO-Detector
4 comments
I linked to the GitHub. The project website is at http://ufoid.net/
cool idea, but the majority of the videos on the "videos" tab make me immediately wary of how useful this will be.
There's one guy (Mike_Egg) who owns the majority of the videos on the page. That in itself is suspicious, since if they were so frequent over a single location, we wouldn't need something like this to catch them. The videos are also VERY sloppily edited, though. Meaning they are feeding fake data to this system and having it flag artifacts they shopped in.
its a cool idea, but I'd maybe put some policing on your user-submitted videos to weed out the obvious fakes if you want to gain credibility that this system works
There's one guy (Mike_Egg) who owns the majority of the videos on the page. That in itself is suspicious, since if they were so frequent over a single location, we wouldn't need something like this to catch them. The videos are also VERY sloppily edited, though. Meaning they are feeding fake data to this system and having it flag artifacts they shopped in.
its a cool idea, but I'd maybe put some policing on your user-submitted videos to weed out the obvious fakes if you want to gain credibility that this system works
I recognise this from the university, most impressive & intriguing final project of the lot - it was praised heavily by the lecturer too.
Bravo on the work, I'll put it to use for sure this summer.
- 2nd year moving to 3rd
Bravo on the work, I'll put it to use for sure this summer.
- 2nd year moving to 3rd
This is really neat! I've long had a curious but skeptical interest in UFOs and I've wanted to see someone do something serious like this.
This kind of approach is the only way we will get actual non-garbage data on this subject. Photos with no pedigree are worthless since anyone with a paint program can fake them. Regular cameras just aren't suited to taking pictures of distant moving objects (try photographing a bird with your phone). Eyewitness testimony is unreliable. People keep claiming to see anomalous things but without real evidence we can't draw any conclusions other than a big "hmm..."
Unfortunately the success of something like this is heavily dependent on the quality of the camera used. Web cams are not going to give you any detail and cheap CCDs are prone to certain kinds of imaging artifacts.
Even worse a lot of consumer hardware either compresses images right away or runs the image through algorithms to smooth, blend, edge-enhance, color-correct, etc. to correct for artifacts present in cheap CCDs. Your "high quality" tiny cell phone camera is actually a mediocre CCD combined with a ton of software to enhance images to make them pleasing to the human eye. For a scientific instrument this is the last thing you want. You want raw data straight from a high quality calibrated sensor.
It would also be worth looking into gigapixel imaging either via gigapixel CCDs or low-artifact compositing algorithms to produce gigapixel images from multiple lower resolution sensors. Here's a few interesting links:
http://project.pan-starrs.ifa.hawaii.edu/public/design-featu...
https://www.theverge.com/2013/2/1/3940898/darpa-gigapixel-dr...
These are going to be pricey but I'm there are enough people interested in this subject for a crowdfunding campaign.
Finally I'd prefer infrared. Due to the second law of thermodynamics anything using a lot of energy must reject waste heat, even if it's using some kind of yet-unknown or exotic means of propulsion. All the candidates I'm aware of for exotic propulsion would have enormous power requirements, lighting up like blinding spotlights in the infrared.
http://www.projectrho.com/public_html/rocket/spacewardetect....
"The Space Shuttle's much weaker main engines could be detected past the orbit of Pluto. The Space Shuttle's manoeuvering thrusters could be seen as far as the asteroid belt. And even a puny ship using ion drive to thrust at a measly 1/1000 of a g could be spotted at one astronomical unit."
This kind of approach is the only way we will get actual non-garbage data on this subject. Photos with no pedigree are worthless since anyone with a paint program can fake them. Regular cameras just aren't suited to taking pictures of distant moving objects (try photographing a bird with your phone). Eyewitness testimony is unreliable. People keep claiming to see anomalous things but without real evidence we can't draw any conclusions other than a big "hmm..."
Unfortunately the success of something like this is heavily dependent on the quality of the camera used. Web cams are not going to give you any detail and cheap CCDs are prone to certain kinds of imaging artifacts.
Even worse a lot of consumer hardware either compresses images right away or runs the image through algorithms to smooth, blend, edge-enhance, color-correct, etc. to correct for artifacts present in cheap CCDs. Your "high quality" tiny cell phone camera is actually a mediocre CCD combined with a ton of software to enhance images to make them pleasing to the human eye. For a scientific instrument this is the last thing you want. You want raw data straight from a high quality calibrated sensor.
It would also be worth looking into gigapixel imaging either via gigapixel CCDs or low-artifact compositing algorithms to produce gigapixel images from multiple lower resolution sensors. Here's a few interesting links:
http://project.pan-starrs.ifa.hawaii.edu/public/design-featu...
https://www.theverge.com/2013/2/1/3940898/darpa-gigapixel-dr...
These are going to be pricey but I'm there are enough people interested in this subject for a crowdfunding campaign.
Finally I'd prefer infrared. Due to the second law of thermodynamics anything using a lot of energy must reject waste heat, even if it's using some kind of yet-unknown or exotic means of propulsion. All the candidates I'm aware of for exotic propulsion would have enormous power requirements, lighting up like blinding spotlights in the infrared.
http://www.projectrho.com/public_html/rocket/spacewardetect....
"The Space Shuttle's much weaker main engines could be detected past the orbit of Pluto. The Space Shuttle's manoeuvering thrusters could be seen as far as the asteroid belt. And even a puny ship using ion drive to thrust at a measly 1/1000 of a g could be spotted at one astronomical unit."
I don't think infrared would be good a idea, I saw an UFO a long time ago passing by close to my home in Peru, it stood still in the sky for a moment and then it flew really fast and lost track of it, what struck me was that it was totally silent, even stealth jets make a noise, the UFO movement was also graceful it didn't seem to be fighting gravity, I think they don't use propulsion, they have some way to manipulate gravity.
Unless they've found a way around the second law they would still have to reject waste heat. The second law is probably one of the very best scientific laws in terms of having been solidly and repeatedly proven.
The most efficient generators known are certain kinds of batteries and fuel cells at over 90 percent conversion. It's conceivable that some kind of exotic betavoltaic nuclear process like (I think) boron fusion could beat this by basically making electrons directly.
Still the second law says nothing can be 100%.
Let's assume 95%, which would be amazing. If you had a propulsion system that required 10 megawatts of power, you'd be rejecting 500 kilowatts of heat at 95% efficiency.
So that's a 500 kilowatt infrared light bulb!
10 megawatts is probably conservative. The only hypothetical candidates for exotic propulsion I know from physics have outrageous energy requirements. Even assuming many breakthroughs I would still assume them to be high.
I can see a 25 watt light bulb at some distance at night.
If (the positive hypothesis) what you saw really was artificial and really was maneuvering like that, it would have been very bright in the infrared.
I do think we can assume that anything visiting us is inhabiting the same universe and is subject to its laws. It's the only sane assumption. The alternative is things like "we are living in a simulation" and then anything goes so why bother.
The most efficient generators known are certain kinds of batteries and fuel cells at over 90 percent conversion. It's conceivable that some kind of exotic betavoltaic nuclear process like (I think) boron fusion could beat this by basically making electrons directly.
Still the second law says nothing can be 100%.
Let's assume 95%, which would be amazing. If you had a propulsion system that required 10 megawatts of power, you'd be rejecting 500 kilowatts of heat at 95% efficiency.
So that's a 500 kilowatt infrared light bulb!
10 megawatts is probably conservative. The only hypothetical candidates for exotic propulsion I know from physics have outrageous energy requirements. Even assuming many breakthroughs I would still assume them to be high.
I can see a 25 watt light bulb at some distance at night.
If (the positive hypothesis) what you saw really was artificial and really was maneuvering like that, it would have been very bright in the infrared.
I do think we can assume that anything visiting us is inhabiting the same universe and is subject to its laws. It's the only sane assumption. The alternative is things like "we are living in a simulation" and then anything goes so why bother.