Some advice for "camera detection"

Hello again!

One part of my current project involves some kind of "surveillance awareness", so I'd like to implement a camera detector.

Since this is for an artistic project, it doesn't need to be EXTREMELY precise or accurate...

So I've been thinking about possibilities to detect the presence of a camera. So far I've come with two ideas:

1) Use en EMF detector with a filter for the frequencies of the CCDs, which are more or less the same in the short wave band spectrum (from 15 to 20 MHz I suppose) and then use the amplitude (probably measured via proximity to the detector) to trigger something out of Arduino (blink an LED, for instance. This doesn't matter for now)

2) Use a visual-oriented detection: if you shoot a beam of red light in a room while observing it through a red color filter, the CCDs/lenses tend to blink in response (there are a few commercial gadgets that do exactly this). But this would get complicated, since I would have to calibrate a color sensor precisely to retrieve accurate data, i.e. distinguish between different red brightness (the filter and the blinking).

one alternative to solution 2 would be without any Arduino board, just by hooking up a webcam to a computer and doing the point detection via Processing or PD/GEM. This would be my last try, since I'd like to keep things disconnected from a computer.

so far the number one, although quite tricky, would be the best possibility. I'm trying to keep the complexity to a minimum in this project.

would you have any other interesting and/or more feasible suggestions?

[quote author=Pedro Oliveira link=topic=87920.msg660333#msg660333 date=1327068213] I'm trying to keep the complexity to a minimum in this project. [/quote]

I would be surprised if any of those techniques was feasible without a substantial research team behind you, and I think it is vanishingly unlikely that you will get anything working within the scope of an artistic project. In any case, the processing involved would be way outside anything you could expect to achieve with an Arduino.

I don't know what type of project you have in mind, how much time, technical skill or time you're able to spend on it, but one possible approach would be to turn your project on its head and create something that does surveillance. For example, some sort of facial recognition controlling personalised advertising. (I know it's not a new concept but it's something that is not commonly available yet but seems technically feasible if you can get facial recognition and discrimination working. This is not an Arduino project, though.

PeterH: I would be surprised if any of those techniques was feasible without a substantial research team behind you, and I think it is vanishingly unlikely that you will get anything working within the scope of an artistic project. In any case, the processing involved would be way outside anything you could expect to achieve with an Arduino.

Hej,

thanks for the answer, although I feel that maybe I didn't make myself clear.

Like I said, I'm aware of the technical implications of doing something really precise and effective, but this is not my goal. I'm not trying to create a spy-detector that would function anywhere but rather a really specific situation which I'd like to retrieve information from. Therefore, such "situation" would be according to the "detection technique" I'd use, for optimal results. So, in a sense, I'm not creating anything that is even close to a commercial-like solution or in any way related to personalized advertisement as you suggested, but rather a really specific set up :)

I'm still thinking about the color detection. I was thinking whether it is possible to calibrate a color sensor (the one from Sparkfun, for instance) to react to only one color and its different brightnesses, in this way I could calibrate the sensor to "see" red as the environment color and then detect certain levels of brightness (the blinking spots you get from the cameras)... a possibility? Or maybe even use a Light sensor like the TSL230R and the same kind of calibration ("teaching" the sensor that red is the environment color - the filter - and then reacting to certain degrees of brightness)

Or... - but then I'd need more advice - any sort of "blob detection" with Arduino? Or is it too much for the board?

Thanks!

Detecting the CCD by it's RF radiation I doubt is technically feasible, at least it would require a substantial amount of R&D.

Illuminating the room in exactly the right colour and then looking for a response (reflection?) from the CCD sounds credible but I have no idea whether that would actually work in practice. If it did work, you would need to use image processing techniques to pick out the return light from the room's image, and the Arduino has nothing like the memory and processing power required for that sort of thing. You would need a PC or similar to do this. Even on the right platform, I think it would be challenging to get it working at all.

Since this is for an artistic project, it doesn't need to be EXTREMELY precise or accurate...

Its just not going to happen. Artistic project request are fairly frequent, but the problem is usually founded in the fact that the requesting person is majoring in art and not engineering and doesn't understand the technical difficulties of the project. I've seen request for a trombone playing robotic arm for a music class project, and request for a robotic arm connected to a camera that paints a person's portrait. Cool but increadibly difficult.

Did you find any solution i'm also working on camera detection Please reply ASAP