I've been reading the IET journal article on infra-red thermal imaging cameras, and their medical applications. It mentions that a thermal imaging camera costs about £20,000 (about $40,000?)! Now, my question is this: why so expensive? Is it something to do woth the low volume of sales? Or is there some part or parts that really puts the price up?
Way back in the 1970s, I saw a thermal imaging camera that could show variations in body temperature. It showed my glasses (actual glass lenses back then) as black, opaque to IR. The lens in the camera was made of germanium -- yes, a shiny metal lens! Is that the reason for the high cost? Do modern IR cameras still have germanium lenses?
Hope somebody who's more familiar with IR cameras can answer this! Would be great to make a thermal imager, but it looks like a tricky problem if glass lenses can't be used.
The microbolometer in a thermal camera is incredibly difficult to manufacture, but not $40,000 difficult. The trick is that since they're seen as high-end devices, any sensors with flaws are discarded, instead of built into cheap cameras. If they were a consumer-grade item, more defects would be acceptable and manufacturing would get cheaper. It's a chicken-and-egg problem.
Anyway, prices are coming down, you can get a very nice thermal camera for $5,000 or so now. Do some research, if you come up with a DIY method I'd be interested!
Myself248 is essentially correct regarding high-cost medical thermal imaging cameras. Low volumes, no defect tolerance, very high quantities of FDA red tape to qualify the device, insurance against lawsuits if it fails to see something, etc. all lead up to a very high end-user price.
What is your application? If you just want to see "hot spots" and don't need the "false color" image, you can take the IR filter off of a B&W video camera and "see" IR. Google IR and B&W camera and you'll find oodles of links. Some camera sensors are more sensitive in the IR range than others, so look around to see what others have played with and had good results with. You may even be able to find a filter that blocks everything except IR, which would effectively increase the sensitivity of the sensor - it won't be blinded by bright lights, just the IR portion of that light. To do a quick check to see if a camera can see IR, flash a remote control into the lens while you watch the monitor output. If you can see the flash, the sensor can see IR.
If you're looking at medical applications, that's a whole other ballgame. I work in R&D designing cardio and neuro stimulation leads and adapters, and the amount of testing and red tape we need to present a case to the FDA that our device is 1) safe, 2) effective 3) better than the currently available stuff is quite expensive. I've got a lead design sitting in a tester that's over 270 million flex cycles, and it has to reach 400 million before it is acceptable. at 15hz, that's over a year of continuous testing. The tester, a Bose Enduratec (and you thought they just built speakers!) was not cheap, and our time to administer the test is not cheap either.
If you want a case study for why the medical products industry is and should be paranoid about testing, and why we charge so much (because that testing is expensive) Google the Therac-25. A software fault caused at least 6 known extreme overdoses of radiation therapy. A thermal imaging camera isn't in the same class - it does not emit stuff at the patient - but the idea is the same, in that if it harms or fails to detect harm to a patient the manufacturer may be held liable.
cajunfj40 sez: "If you just want to see "hot spots" and don't need the "false color" image, you can take the IR filter off of a B&W video camera and "see" IR. Google IR and B&W camera and you'll find oodles of links."
This won't help the original poster if he's truly looking for IR thermal. Thermal imaging uses long wave IR. Normal camera sensors are sensitive to short wave IR, but this won't show thermal "hot spots".
c0redump: I saw an article on the web a while ago about a guy who made his own thermal imaging camera for under $100. He used the polymer sensors from PIR detectors - it was *very* low-res though. Unfortunately I couldn't find the article with a quick google, but it may still be out there.
Not sure if this helps, but I used to use a thermal imaging camera for trouble shooting industrial wiring. It was a camera made by Fluke the same folks that make multimeters. It worked great for finding loose connections, bad breakers, failing components. When you took images of people you could see quite a bit of difference in surface temperatures and it was a couple grand. However, Fluke products are a bit pricey and it had software for comparing images etc. I have to believe that one could build it cheaper.
You can make a simple thermal imaging camera with $50 or maybe even less. it doesn't have an LCD though, it's kinda like a "mechanical TV". I'll tell you more later.
I'm getting into the home energy auditing business. I just bought a Flir/Extech i5 camera for the job. Its about $3000. It has an 80x80 pixel array, which isn't super high resolution, but is good enough to see the swath of cold left by air leaking around doors, windows, and sill plates.
Phil Hobbs' project uses ferroelectric/pyroelectric polyvinylidene PDVF film, the same sensor in IR motion burglar alarms and security floodlights.. To avoid buying 1000 gigohm resistors and to improve diode muxed switching, he uses light-biased red LEDs as the diodes.
Note that pyro plastic film only responds to temperature *changes.* Either it only can see moving sources (a thermal "frog's eye" effect,) or the camera needs a rotating chopper wheel.
I don't think he mentions the lens. Pinhole? Polyethelene fresnel lens? Telescope mirror?
Seems like something useful could be created with a single pixel of IR-sensitive material and a mechanical mirror/voice-coil scanning arrangement. Sensitivity would be horrible, but a tripod might allow one to produce a fairly high resolution image given enough time - seconds, minutes?. There would be some averaging as a result of the exposure time.
Yet another idea not suitable for the original poster, but a thought nonetheless.
If I were going to tackle this, I'd start with one of the inexpensive IR Thermometers that Harbor Freight is selling. Figure out how to get a signal out of the thing, maybe put a pinhole over the sensor to tighten up the dot size, and then build some sort of mechanical scanner. The scan speed would depend on the responsiveness of the sensor.
One rotor simple-radio-controlled or telemetry-controlled helicopters are commercially available. These Search & Rescue and Security helicopters usually are equipped with a very good digital camera. Sadly though, far fewer UAV helicopters currently have IR Infra-red heat imaging cameras onboard.
[ quote from a recent makezine commentator ] - "The trick is that since they're seen as high-end devices, any sensors with flaws are discarded, instead of built into cheap cameras. If they were a consumer-grade item, more defects would be acceptable and manufacturing would get cheaper. It's a chicken-and-egg problem. Anyway, prices are coming down, you can get a very nice thermal camera for $5,000 or so now. Do some research, if you come up with a DIY method I'd be interested!" [ End Quote ]
So, on that note [ immediately above ], why not design, build and offer for sale very reliable, foolproof radio or telemetry operated **** one rotor UAV camera helicopters **** for Search & Rescue and Security purposes? I cannot yet find any Extended Range versions with a fuel cell and compressed, metal-hydride or liquid hydrogen options that are for sale to us technologically hampered civilians.
Once done, you'll sell many THOUSANDS worldwide. Afterwards, maybe then you'll have enough of a customer base so that you can contact all your clients & offer a " consumer " quality version of the IR Infra-red heat imaging camera .... for Search & Rescue and Security use.
Remember, limited range [ paltry poor flight duration ] is currently the main limiting factor in [ common civilian ] UAV Camera Drone technology today.
Doesn't have to be that way though .......
Me, I want one UAV at " elevation " immediately above me with a telemetry operated telephoto lense onboard it that's pointed at the other Extended Range UAV Helicopter up to 10 miles downrange. This way, a constant "line of sight " is maintained between the UAV hovering almost immediately above me and the one that's out there up to 10 miles radius from my vehicle.
With this very interesting strategy, I'm guessing that I'm technically " legal " with current FCC rules on using unpermitted Unmanned Aerial Vehicles. Right? The rule says that a " line of sight " must be maintained at all times.
Any questions? Find a very rare copy of "Electric Car" magazine from 1995. Alan Cocconi at AC Propulsion in San Dimas, California was featured in this low-circulation ' market test ' magazine about electric vehicles. Call www.acpropulsion.com . Read all about the ' So Long ' UAV and Alan's experience with UAV camera drones in the 1980's & 1990's.
While you're at it, pick up a 248 [ maximum ] horsepower Electric Vehicle Motor with the necessary PWM / MOSFET power control electronics. Price? A mere $27,000. Best of the best for sure!
This powerful electric vehicle motor is currently installed in the www.acpropulsion.com T-Zero sports car and the new Tesla electric sports car at www.telsamotors.com
Bring me onboard in your company if I can spill any more profitable beans.
Employment is very scarce for those with less than a proper 4-year degree right now ....
Heimann makes 2D thermopile arrays. The 8x8 version with the sensor only is $300. There are more advanced versions that have integrated boards with an ethernet adapter built in and windows software to interpret the stream.
I don't know if this is out of topic or not, but there are thermal imaging cameras that are used for surveillance and security purposes in general. I don't know their prices but I can say that they are high quality ones. The site is this one: http://www.controp.com/category/thermal-imaging-cameras
With the intention of making a cheap night vision camera i bought a rocketfish RF-HDWEB10 1080p USB camera, after removing two screws and removing the faceplate i had access to the AF element, removed the focus lense(s?) and on the back of that focus element was an IR filter, wich i chipped off, then i replaced the focus group and connected the camera to my computer, and found that my remote controls now lit up the whole room, but i also noticed that my cousin's soldiering iron was glowing bright white, so with all 4 of us extremely ecstatic with the possibilites we used an IR laser and got a constant reading of 280F with the lights out the lowest temperature that was visible was about 150F, but determined to build our own thermal camera for under $21USD we found that glass seems to block the small amounts of thermal IR coming off of skin, so i removed the focus element and plugged it in without any glass in front of the sensor, and sure enough waving my hand in front of the sensor at a few feet away made the computer's( in the other room monitoring it) screen flash!!!!! so what i need now is a glassless focus group, we're trying to find a mirror system to focus onto the sensor but any suggestions would be greatly appreciated, i also found some lenses made for thermal camera's but i could not find a price and i am wanting to keep the cost under $50, it doesnt need to be "less than .0001F fault" really 10F of error would be amazing
here are the lenses http://www.iiviinfrared.com/ir-thermal-imaging-optics/lenses short demo of what the camera does http://www.mediafire.com/?7asoa1aa389jv the light that you see in the picture and video is an IR LED flashlight and the thing i am holding and that the camera is seeing thermal on is a blow dryer, temp'd at 230F (might have been colder as i was outside on a cold day) I also modified an olympus SP800UZ in the same manor(though that took alot of effort) and i found that the "glass blocks thermal IR" statement to be mostly true because even though the two camera's have very similar sensors(the olympus is a bit bigger and more sensitive) the olympus has a verry hard time seeing any thermal, the only thing it would see was my hot exhaust manifold at nearly 550F and it was barely visible, the main reason(i think) is because there are 9 glass elements in the olympus(some of them are very large too) where as the usb camera has only 2(i think)
As you've found, glass [1] is mostly opaque to infrared [2].
[1] Be a little careful, since the word "glass" covers a wide range of materials, with widely varying compositions and characteristics.
[2] Also, be a little careful with the word "infrared", since this covers a huge band of wavelengths, everything from near-infrared, which is just below the optical band, down to far-infrared, which is quite removed from the optical band, and is more like microwaves.
Thus, most cameras designed for infrared viewing don't use glass lenses. There are several materials which can be used, including fused quartz (usually the kind produced in an electric oven, to minimize the concentration of hydroxyl radicals, which have a strong absorption in the infrared range). Germanium can also be used, but Germanium is quite expensive, with somewhat wild price fluctuations. Even some plastics may work. The trick is that the material has to be transparent to infrared in the range of interest, and that it have a reasonably constant index of refraction over that range of interest.
Of course, it's possible to use a reflecting design to focus an image. This technique is quite common for telescopes, although cameras usually don't usually require this level of complexity (although there are some reflecting "lenses" for cameras). But, this is a way of getting around having to send the light through a lens. The trick with such a design is to choose a first surface coating for the mirror which reflects in the infrared spectrum of interest.
The other trick for building an infrared camera is to choose an image sensor which responds to the wavelength of interest. Most common digital cameras use Silicon based image sensors, which, while sensitive reasonably far into the infrared region, are limited. This primarily has to do with the band-gap energy of Silicon, and photons with energy below the band-gap energy aren't energetic enough to cause electron-hole production in the sensor. Lower band-gap energy materials are available (Germanium, Lead Sulfide, etc.), but tend not to be mass produced (well, at least in the same quantities that Silicon based sensors are), which leads to an economy of scale problem.
Another problem for far-infrared sensors is the temperature of the sensor itself masking the incoming radiation it's trying to measure. As a result, some far infrared sensors are cooled to cryogenic temperatures to prevent the sensor from masking the signal. This cooling can be accomplished by a liquid Helium or liquid Nitrogen bath (e.g., evaporative cooling), or with something such as a Peltier cell cooler.