Ryan Spicer
August 30th, 2005, 07:27 PM
I'm working on some interesting stuff this semester involving using a live video image of a performer on a proscenium stage, analyzed with a computer running eyesweb, to generate realtime visualizations to be projected behind the performer.
My advisor and I are looking into ways to avoid the analysis quality being affected by dynamic lighting conditions, and one of the possible solutions we've decided to research is looking for a camera that captures only IR light, and illuminating the stage with IR lamps that remain at fixed intensity and geometry throughout the piece. I've googled a bit, and the cameras seem to break down into "0 Lux" security cameras that work as visible-light + IR cameras (not suitable becuase I want to ignore visible light entirely), and $20,000 heat-sensative cameras (not suitable because they cost an arm and a leg). I've also seen a few filters that claim to block visible light, leaving an IR-sensative camera with IR light only. Are these remotely effective?
Keep in mind that the captured image will only be used by a computer, which will be performing a basic difference-key, then running a series of algorithms from eyesweb on each frame, so quality isn't an extreme concern, as long as its a low-noise image. I don't mind an analog solution as long as it has rca/bnc/svideo outs so I can run it through a transcoder box.
Any thoughts you folks can offer would be greatly appreciated! I know this is rather off-topic for DVinfo, but this is the most relevant group I know of.
Cheers,
Ryan Spicer
My advisor and I are looking into ways to avoid the analysis quality being affected by dynamic lighting conditions, and one of the possible solutions we've decided to research is looking for a camera that captures only IR light, and illuminating the stage with IR lamps that remain at fixed intensity and geometry throughout the piece. I've googled a bit, and the cameras seem to break down into "0 Lux" security cameras that work as visible-light + IR cameras (not suitable becuase I want to ignore visible light entirely), and $20,000 heat-sensative cameras (not suitable because they cost an arm and a leg). I've also seen a few filters that claim to block visible light, leaving an IR-sensative camera with IR light only. Are these remotely effective?
Keep in mind that the captured image will only be used by a computer, which will be performing a basic difference-key, then running a series of algorithms from eyesweb on each frame, so quality isn't an extreme concern, as long as its a low-noise image. I don't mind an analog solution as long as it has rca/bnc/svideo outs so I can run it through a transcoder box.
Any thoughts you folks can offer would be greatly appreciated! I know this is rather off-topic for DVinfo, but this is the most relevant group I know of.
Cheers,
Ryan Spicer