21 April 2006

False Color Infrared Photography with a Webcam

John F. Green

False color near-infrared photography was invented by researchers at the Eastman Kodak Company during World War II for the purpose of camouflage detection. The method was based on the fact that living vegetation reflects near infrared sunlight much better than artificial vegetation and camouflage paints and netting. It has since been shown to have many other uses.

Near-infrared photography is widely used for aerial photography, because it eliminates the blue haze that is characteristic of the atmosphere. It is often capable of discriminating between healthy and diseased or stressed vegetation. It can detect certain types of water pollution. It is sometimes used for artistic effects by photographers.

Special films, filters, and handling techniques must be used with color infrared photography. For example, color infrared film must be stored in a refrigerator or freezer, loaded and unloaded in total darkness, and developed within two days after exposure. Thanks to electronic imaging technology, all this can now be avoided.

Here I will describe a procedure that uses an easily modified Webcam and an inexpensive infrared filter that closely replicate the results provided with false color infrared film. Please note that we are dealing here with near-infrared radiation, not with far- or thermal infrared which is used for nighttime surveillance, thermal imaging, etc. I mention this because I find that people often seem confused on that point. A webcam won't help you find a fugitive in the dark unless you are illuminating the scene with short-wave infrared light.

Figure 1. Flowers in false color infrared.

R ADJACENT TO FIG. 1 ABOVE.]

Figure 2. The same flowers in Fig. 1 in visible light.

The human eye can detect the optical spectrum from about 400 nanometers (violet) to about 700 nanometers (red). Our visual system perceives this range of light wave frequencies as a continuous rainbow of colors. Charge-coupled devices (CCDs) are light sensor arrays that respond from the blue to about 1000 nm in the near-infrared. The range from 700 to 1000 nm falls within the near infrared region of the spectrum.

Because of the high sensitivity of CCD sensors to near-IR, digital camera manufacturers usually place a filter that transmits visible light but blocks near-IR between the lens and the CCD. The absence of such a blocking filter was the basis of the so called X-ray vision modifications to video cameras a few years ago. In order to use a camera for near-IR applications, this filter can with some effort be removed, but this step is usually unnecessary because the filter does not block all the near-IR.

How to Make a Near-Infrared Image

The technique I have developed for making near-IR images is fairly simple. It requires a double exposure, so a tripod or some means of rigidly mounting the camera is required. Having a laptop computer is very helpful for outdoor work. I have used an IBM PC Cam (XVP600) and a Logitech ClickSmart 510 and have received similar results with both. Depending on the specific camera your results may vary.

The procedure is as follows:

1. Compose your scene. For initial experiments, I suggest an outdoor scene with green vegetation and bright sunlight.

2. Make the first exposure.

3. Without moving the camera, place an IR filter over the lens and repeat the exposure. I usually let the camera software determine the exposure, which it generally does tolerably well.

4. Once you have the pictures, load them into Adobe Photoshop or any other photo editing program that allows splitting an image into separate color bands. Create a new, blank 640 x 480 RGB image (or the size your camera uses).

5. Next, go back to the image taken with the IR filter. In the Photoshop menu, select Channels. Select the red channel. On the Select menu, click on Select All. Copy it to the clipboard. Select the new image, select the red channel and do a paste. Now select the image taken without the filter. Again select the red channel, copy it to the clipboard and paste it into the green channel of the new image. Now repeat the process with the green channel of the normal image, this time into the blue channel of the new image. In other words:

a. IR red channel to new red channel.

b. Normal red channel to new green channel.

c. Normal green channel to new blue channel.

6. Now when you click on the new image RGB channel, you should see something resembling a false color infrared image of the scene. This is similar to the way false color infrared film works though it uses a CYMK (Cyan-Yellow-Magenta-Black) scheme while here we are using RGB (Red-Green-Blue). At this point you may want to adjust the colors slightly with the adjust feature under the image menu of Photoshop.

.]

Figure 3. Scene showing a typical near-IR false color view of of healthy vegetation (red).


Exposure Time and Near-IR Filters

You will find that a considerably longer exposure (as much as several hundred times) is needed for the picture with the IR filter in place.

Infrared light will come to a focus slightly before the image plane compared to visible light, but since most webcams have such small lens apertures, this can usually be ignored unless you are doing close-up work. In this case experimentation is probably your best bet. Some experienced photographers suggest focusing 10% short.

I use a simple, cheap IR filter from Edmund Industrial Optics (Stock No. H43-948). This filter is not intended for imaging applications, but at 640x480 pixel resolution it is more than adequate. Many other kinds of filters are also available from this supplier.

An inexpensive near-IR filter, I am told, is a frame of unexposed but developed positive transparency (slide) film. If one is interested in better quality, there are a number of near-IR filters made for this purpose available from photography suppliers. I won't go into a lot of detail about filters because many different filters are available and the correct choice depends heavily upon the particular situation.

I mentioned above that one can remove the camera's IR blocking filter. In this case, however, you will have to find some way of replacing it or a similar filter somewhere in the optical path while making the normal exposure. The filter removed from the camera will in most cases be too small to be conveniently used in front of the lens. Removing it and using a replacement in front of the lens, however, will considerably improve your results, because by leaving it in place you cause the CCD to sum some visible light with the infrared.

Figure 4. The image on the left was taken with the cameras infrared blocking filter removed while that on the right was taken with the filter in place. Note how bright the vegetation is and how dark the water is on the left.

Going Further

There is no reason this procedure cannot be used with more elaborate digital cameras, though I would be reluctant to disassemble my digital SLR. Some digital photographers have been venturing into the infrared (and even ultraviolet) realm for years with spectacular artistic results. Digital cameras have opened new pathways into a fascinating and often beautiful world beyond the range of the eye to those willing to experiment and perhaps break a rule or two now and then.

Plugging infrared photography into your browser search line will bring up a lot of reading material. For example, here are a couple of interesting web sites:

http://folk.uio.no/gisle/photo/ir.html

http://www.naturfotograf.com/uvstart.html


   
Copyright 2005 by Society for Amateur Scientists