Ideas for a UFO Sensor

[marduk]

I've had this idea kicking around in my head for some time.

I have a few nest indoor and outdoor cams for security, and to catch pics of wildlife, and I often also scan through the night shots for anything anomalous in the sky.

I've been noodling about repurposing my outdoor cam to purely point upward to the sky. I think it should probably be somewhere that doesn't point at the sun during the day so it doesn't damage the ccd. The software automatically captures movement, so that would be enough to start.

But one could easily add an old iPhone or iPod to the mix - they actually have a built in magnometer to measure your location and direction. You can get an app to monitor those fields. Not sure if it would be sensitive enough or not, but it would be something. A second iPhone could act as a gravimeter (using the built in accelerometer).

One could even contemplate using something I think Wilbert Smith came up with -- some kind of spacial sensor that ended up just being a piece of material being brought just to the point where it will tear but not quite. I remember reading how he surmised it could make a good sensor for UFOs for some reason. I can't find the reference.

A second nest cam could be used to monitor the two iPhones and the Smith sensor, while the first watches the sky.

The range would be limited to where you can run power and connect to wifi, but it could make for a pretty cheap and purposeful UFO sensor.

Thoughts?

 

[Trevor Wozny]

Good Morning Marduk,

Awesome ideas. Always love when the dendrons move. If your consideration is strict with optical (our little infinitesimal sliver of EM) then it's gonna cost ya. Commercial grade cameras are NOT up to the task. What you are going to want is a very high-end camera which allows you dump its output RAW.None of the common avi, mpg, whatever. All will completely nullify any evidentiary weight. The reasoning being that those compress not document. By a bunch of averaging, normalizing, all that stuff...what ends up on your memory stick is Art. Not Evidence. You can never enhance something by changing it. Right? You need to get as close to what those chips sense if anything meaningful can be hypothesized. It gets worse, you're going to need at least 3 of them. All pointing up with calculated overlapping fields of view (all cameras can see all other cameras) These parallax views are essential if things like distance, azimuth, etc. A photo alone will never provide any substantial evidence. You can never know a phenomenon, you can only quantize what it does. the abstractive distance is insurmountable. Some day we will have a logically consistent hypothesis of what it does...we may never know what it is. Sorry if that seems like a downer..it really is not. Peace.

 

[marduk]

I'm thinking of going with what I have that I can repurpose instead of going all state of the art. Grassroots kinda deal.

Because the thing is, I could set this up and run it for a year and not catch anything. That's the most likely scenario.

But if a bunch of us throw something like that in our backyard, that might be something.

 

[Trevor Wozny]

Sir, the effort, and diligence is its own reward. new ideas can change the world... really. Don't ever listen to anybody that tells you can't do something. realize...what do they know.

 

[DROBNJAK]

You can do even better. There is a sensor for Arduino that can track objects. Like when you use digital camera and it can track faces. This arduino sensor is relatively easy to program and you can train it to "box" high contrast objects, like planes against blue sky or clouds. Than you link all that to servo motors and you can have a sensor picking up every flying object, say above the horizon. Than you just take videos or photos.

 

[DROBNJAK]

For anybody who wants to do this as a project, here is the actual product:

CMU Cam v5 (maybe there is a newer version).

In this application, camera is more or less irrelevant, so just any Arduino camera from eBay won't do the trick. Important thing is that software that will do object recognition. CMU CAM v5.0 has that software built in on a separate CPU, so its super fast.