Thread Tools
This thread is privately moderated by otlski, who may elect to delete unwanted replies.
Mar 22, 2013, 09:00 PM
checkout my Blog
otlski's Avatar

IR camera options

To start, I see a lot of misinformation regarding night vision, infrared cameras, and various CCD / CMOS technologies. I have made several posts throughout several forums. The following is a compilation of some of them.

A few months ago, i was approached by an Open Pilot member inquiring about IR camera choices. Below was my reply

I've looked at FLIR cameras for my own use as well. Having one would keep me familar with the problems our customers face. The model I keep coming back to is the FLIR PathfindIR. It is available in 30 fps with a NTSC output (if you are a US citizen). The output is a monochrome image (grayscale). They are 320 horizontal by 240 vertical resolution; for low-end FLIR, that is decent resolution. The FOV might be an issue at 36 X 27-deg. Because of the FOV, you would have to deal with the "tunnel vision" as well as optically amplified affects of airframe vibrations whether with or without a gimbal. These cameras were used by Cadillac and BMW in their night-driving option package.
you'll likely need their proprietary cable connectors to make hook-up easy.

I also seem to recall seeing that the FLIR Firstmate series has a NTSC / PAL output. They offer differing resolutions and FOVs. All outputs are grayscale.

A few weeks ago I picked up a FLIR I7 hand-held for various purposes: home cold air infiltration, attic critter detection, wood stove troubleshooting, general mechanical troubleshooting, etc. One of my long term goals is to gather hardware to do live image fusion with color visible light, near IR, and long wave IR (FLIR). I picture combining information from any of the two sources, perhaps feeding luminance from one source with chroma from the second source. I think doing so would mitigate the affect of the inherently low resolution of the affordable FLIR cameras.

I wish you good luck with your project. I'm happy to help more if I can. There is a guy on RC Groups CenTexFlyer who I think has fixed wing SAR experience. He might be a good resource to talk about resolution, FOV, altitude, and detectability. If someone like him says that 36 X 27-deg is usable then the PathfindIR might be what you want.

And now from a post to Open Pilot Forums regarding cameras for night flying

Let me come at this from a different perspective. My comments come from my other pursuit of video astronomy, and having a fair amount of experience with bleeding edge cameras and gimbals. It is worth clearing up some nomenclature to start. There are significant differences between the terms night-vision, IR capable camera ($50-$150 price point), thermal IR, and low light CCD and CMOS sensors. You see, I could not even write the previous sentence without being descriptive.

A term such as night-vision is often applied to all the technologies mention above and often is by misused by marketing people. It really should be reserved for photomultiplier tubes and for EMCCDs’ Photomultiplier tube devices are large, heavy, and fragile, not suitable for us. EMCCDs are where onboard the CCD chip, in the area where the pixel charges are shifted off the chip, the charge shift registers have charge avalanche multiplication gain. This is relatively new technology and expensive, but is effective and orders of magnitude better than after CCD readout op-amp voltage amplification. The images produced are somewhat noisy but are usable. I have not seen any EMCCD based cameras that have small package size. Whether phototube or EMCCD, some light is required for these to work.

Thermal IR (FLIR) is a technology where the focal plane array (FPA) a.k.a. the image sensor is manufactured from completely different materials than CCD or CMOS sensors. This is necessary because the thermal spectrum is far removed from the visual spectrum and CCD/CMOS silicon-based chips are not even close to being able to respond to heat. These FLIR cameras operated in what is known as the long wave IR part of the spectrum where the signature of a human can be seen in complete darkness. Note: the military uses mid wave IR for plume detection as well. FLIR will work in total darkness.

Silicon based FPA; visable spectrum CCD and CMOS. Must have some light to function unless aux light is provided. Starting with CMOS: although great advanced continue to happen with CMOS technology where the best of CMOS FPAs will outperform some CCDs, the best of CCDs still are the most sensitive. CMOS APS (trickle down technology from aerospace applications) and CMOS FPAs such as the SONY Exmore continue to push the boundaries in their low-light performance. It has been predicted for the last 10 or 15 years that CMOS would match CCDs in low-light performance but it is not there yet. Now for CCDs: there are many variants and important considerations. It is worth considering these points because price and availability will certainly make this the choice for most of us. First, if you are primarily going to apply this to night-ops, look for a monochrome (black and white) sensor. Color sensors have color filters arranged in a Bayer pattern over each pixels site. These filters reduce the photon flux striking each pixel site resulting in less electrons being liberated (less charge and less signal), and thus, less sensitivity. Furthermore, during night-ops there is no appreciable color saturation in the scene. Colors are very muted. The lack of these Bayer filters make the monochrome sensor twice as sensitive as their color counterparts.

When looking at monochrome sensors look for sensors with response into the IR range. Sounds contradictory with what I’ve already said, does it? It is not, here’s why. The IR that CCD cameras are sensitive is called the near IR band (NIR). NIR is an adjacent neighbor to the visible spectrum. NIR is nowhere near the frequency of thermal IR. They are two different animals. Again NIR is not heat; light is still stimulus. How these NIR CCDs give you an advantage is that they respond well in the visible range and then add to the incoming flux by including “extra” flux from the NIR.

Most monochrome CCD cameras should come without an IR cutoff filter, most color CCD cameras will have one. In either case this filter should be removed to maximize the low-light performance. Once again, the presence of the filter will cut down on the total flux. Removing the filter on either color or monochrome cameras will soften the image. A clear focus point will be lost because the NIR and visible wavelengths will want to focus at different points. Color cameras will also lose their color balance and daylight images will look wrong colored. Both the sharpness loss and the color balance are unfortunate trade offs caused by removing the IR cutoff filter but it is worth it in the sensitivity gain (perhaps another 20%). Don’t worry about color though because you really want monochrome anyway.

Pixel size and resolution. Okay, imagine an FPA that is ” diagonal which is typical for a CCD. Imagine one with 1.4 Megapixel and one with 1.1 Megapixel. All things the same, the lower resolution version will have larger pixel sites (fewer of them in the same area). Larger pixel sites generally mean correspondingly greater sensitivity. Be careful to make this comparison between like diagonal sizes using the same technology from one manufacturer. Pixel site size is referenced in the Sony (see where I’m heading with this) CCD datasheets.

Lenses - SWAP – size, weight, power. SWAP is a whole discussion on its own. Suffice to say the lower the payload mass (FPV camera) the longer your flight times should if you’ve factored rpm efficiency into the design. However, I’m going to suggest you look at one specific area to sacrifice flight time for night op usability; the camera lens. I would look at cameras with c-mount lenses over the microlens cameras. Simply put, light gathering power is directly related to lens aperture. Again, there are many variable including glass qualities, number of glass surfaces, coatings, etc, but the principal is valid. The light gathering power is proportional to the square of the lens diameter. A typical c-mount lens is 3 to 7 times the diameter of a microlens camera. The difference is significant. Another major factor is focal length. Short focal length, wide angle lenses will be brighter than long focal length lenses. The difference in light flux hitting the CCD is inversely proportional to the square of the focal length, the shorter the better. However, there is a big trade off here; if the FOV is too wide, the ground resolved distance will be too small, nearing useless for FPV at fish-eye extremes.

If I were starting a night ops FPV camera system and FLIR and EMCCDs were out of my price range, I would look for a camera equipped with a monochrome Sony EXView HAD CCD sensor; either the 1/3” or the ” variety. The ExView HAD is mature technology (although newer variants have less noise) and is inexpensive. It features specially doped pixel sites extending the sensitivity into the NIR and enhancing the red and green response (that’s right… even with monochrome we still evaluate the spectral response by color response). The chip features tiny microscopic lenses over each individual pixel to increase the fill factor (less dead space between pixels) further increasing response. I would go for one that accepts C-mount lenses so that the extra sensitivity is not crippled by a grossly inferior microlens. The PC164C is a camera that comes to mind. Obviously once an appropriate lens is added the mass penalty is quite large.


Starlight cameras (Mintron) typically use Sony Exview HAD CCD sensors. And achieve their “starlight” viewing functionality by extending the electronic shutter (exposure time) out to 2 seconds or even up to 8 seconds. So at 2-seconds your frame rate is frame per second (FPS). The FPV you are used to flying with is 30 FPS. My guess is that perhaps 10 fps is the lowest you might want to tolerate before video latency makes it unpleasant. I own and use a Mintron starlight camera might consider it a viable option if it were not so heavy. The previous poster who pointed out the LED illumination scene was disingenuous by showing the results from an 8 second exposure hit the nail on the head. It is an un-useful demonstration for our purposes. That is not to say that auxiliary illumination is not worth exploring because I think it is, but be realistic about the expected results. The results will depend on the proximity to your target and all the above mention factors. It sounds like a mini mag-light appropriately lightened is a good place to start to start experimenting.

To close, I’ve likely put most of you asleep with my diatribe; my opinions are not the be-all, end-all. Far from it. I think that as a topic like this progresses, especially to the build phases, effective solutions using different combinations of technology, with different trade-offs, appropriate for different applications will evolve. This crowd-sourcing aspect of mulitrotors absolutely astounds me. The rate of progress and change from the vast community, the number of successful experiments (and equally useful failed experiments) feeding into that progress is unmatched in industry (where I make my living). Open source and community sharing develop what would take $$$ and years to develop in normal channels. In other words, keep talking and keep experimenting. No amount of theory will replace what eventually works well.

More on IR verses near IR cameras from a post on Multirotor Forums

There seems to be some confusion here. First, A "thermal" camera is absolutely an infrared camera; specifically it can see in the medium or longwave IR. These thermal IR cameras use microbolometer sensor arrays which are not silicon CCD or CMOS sensors. The statement that most CCD sensors are sensitive to IR is somewhat true. It is true in that it can see in what is called the near infrared (NIR). This is the spectrum immediately adjacent to the red color in the visible spectrum, it is not "heat" sensing. NIR is useful for vegetation studies. For clarity of nomenclature I would always call a thermal camera an infrared camera and would never call a modified CCD or CMOS camera an infrared camera. Instead, I would call them NIR cameras because that is what they are. Second, the NIR cameras come with either monochrome or color CCD chips. The IR pass filter reduces the information to a grayscale looking image. Thermal IR cameras come with monochrome microbolometer chip sensors which produce an output voltage equivalent to "grayscale". The camera's on-board processing can let the user interpret the thermal information as monochrome or as several different color palettes. In other words, the thermal IR camera can look either color or B&W.

More on NIR imaging

If a key chain type camera is what you are after, look for a small video camera the uses a Sony ExView HAD CCD sensor. These CCDs have high sensitivity extended well into the near IR spectrum. Cameras equipped with the Black and White versions of CCD will not have an IR cut off filter, the color versions will have one which will have to be removed. Removing the IR cut filter can be easy on some and difficult on others. The basis for my info comes from my other hobby, video astronomy. Try googles that include the terms -
Exview HAD
removable IR
video astronomy

-and see what comes up.

Tech support at Supercircuits may be able to help you out.


The standard Sony chipset and CCD allows shutter times between 1/12000th-sec and 1/60th-sec per field, and two fields make a frame. So when the CCD is staved for photons the exposure will lengthen to the maximum - and yes, as you've identified, this will place a premium on the stability of the tri and the camera mounting. Because of the ccd’s interline architecture interlacing artifacts will be problematic unless vibration is kept to a minimum.

IR cut-off...
The filter that needs to be removed is an IR cut-off filter (not IR pass). IR cut-offs are installed by the manufacturer because the light waves in the near IR, red, green, and blue portions of the spectrum do not focus at the same setting when passed through the vast majority of glass lenses. By installing the filter the image looks sharper but is starved of the IR photon flux. Removing it (as you will) allows the IR response but causes an un-sharp focus.

IR pass filter...
Available from Edmund Optics. This is what you add to retain the now added IR response but block the visible spectrum. With this filter added the image will focus nearly as sharp again and you'll achieve the "contrasty" NIR image you wanted.

I love absolutely Rogue_streak's idea using three cameras. With it you can achieve sort of a tricolor imaging "think Hubble". Adding color to the information display gives the pattern recognition part of our brain something it likes to work with. Lots of potential for finding the needle in the haystack.

The Lumix solution sounds elegant but expensive. It is probably not an interline ccd so vibration will not pose as much of a problem. Naturally, resolution is higher but sensitivity is likely lower due to smaller pixel sites.
Last edited by otlski; Mar 22, 2013 at 09:29 PM.
Sign up now
to remove ads between posts
Jul 03, 2013, 03:04 AM
steadiworkz's Avatar
but can you see UFO's with them ?
Jul 03, 2013, 09:21 PM
checkout my Blog
otlski's Avatar
only if they want to be seen
Jul 04, 2013, 03:40 AM
steadiworkz's Avatar
Haha good answer

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
For Sale Infrared night vision HD camera with 139COMS 700TVL board + IR-CUT+Infrared lamp+lens 123456RC FPV Equipment (FS/W) 1 Mar 23, 2013 11:03 PM
Discussion Best outdoor helicopter with built-in camera & IR night vision? DanReed Electric Heli Talk 12 Mar 07, 2013 12:31 PM
For Sale Wildgame Micro Red IR 4 "Trail Camera" crd Non R/C Items (FS/W) 0 Mar 03, 2013 07:42 AM
Discussion Long range IR control of a camera 0l1v3r Aerial Photography 14 Feb 25, 2013 04:31 AM
Discussion Cameras with IR switches for sale. Talontsi96 Aerial Photography 0 Jan 30, 2013 02:48 PM