Kite aerial photography in the near infra-red and ultraviolet

e-mail John

Photographic Techniques Index

Kite Aerial Thermography

Infra-red KAP

These three photos of Armadale were taken by us in July 2008, with a Pentax Optio E35 modified for use in the infra-red by Dr Mila Pravda.

We have chosen these photos to illustrate how lighting conditions affect the images (Colour removed).

To the left, the image has a uniform lighting of soft sunlight.

Lower left, harsh evening sunlight.

Lower right, strong sunlight on the horizon with the foreground in shadow (similar to the Barbauchlaw Glen photo above) taken from the same position as the photo on the lower left but looking approximately northeast rather than south.

Our first outing with the modified Optio E35 (Larger version).

Power Sled 24 kite, continuous shooting mode, sports setting, ISO 80, f5.4, 1/500sec.

Compare the clarity of these images with the ones here which illustrate the effect of haze.

Our second outing with the modified Optio E35  (Larger version).

Power Sled 24 kite, continuous shooting mode, sports setting, ISO 80, f2.7, 1/400 sec.

Our second outing with the modified Optio E35 (Larger version).

Power Sled 24 kite, continuous shooting mode, sports setting, ISO 80, f2.7, 1/400 sec.

Looking vertically down on archaeology site 1  (Larger version). August 2008 Linlithgow Palace

Looking vertically down on St. Ninian's Chapel, Isle of Bute (3D version).

December 2008

The two photos above were taken in overcast conditions. These light pink images were produced from the original red/deep pink ones (eg here) by using the 'auto levels' function in Photoshop.

Near infra-red can also reveal features in grass which later show up as

parch marks in drier warmer weather. Bristol, April 2009

The two photos above were taken in direct sunlight.

Kinneil Roman Fortlet.  

Near-IR using converted Fuji F30 camera

The position of 1980/81 excavation trenches are visible to the right. April 2011

Composite Photoscan NIR KAP Shot (from 3D) of Rubha Aird a' Mhuile Kelping Bothies, Uist.

Near-IR using converted Fuji F30 camera

Jim Knowles  July 2010


Rufford Abbey 

Near-IR using converted Fuji F30 camera

Jim Knowles  July 2011


Another still from a Photoscan 3D model  

Near-IR using converted Fuji F30 camera      June 2011  

Here is a simple 8.5MB low resolution 3D pdf with some smearing on obscured verticals.


A near infra-red, kite aerial photo of Mallace Avenue, Armadale, taken with a 808#16D HD video camera on the lower end of a wire pendulum (not recommended).

In a strong, gusting wind, this arrangement was very unstable but, once again, images were still extractable.  1 April 2013


Our archaeology group's equipment currently includes the following near IR cameras:

Fuji IS-1 and converted Fuji F30 (x2), Fuji F5600, Pentax Optio E35, Cannon 2200 and 808#16D video cameras

May 2013


Ultraviolet KAP

Ground-based tests with a Fuji S5600, modified for both IR and/or UV (also by Mila), suggest that the camera would be sufficiently sensitive for IR KAP work (using an R72 filter) and also IR and/or UV through a B+W 403 filter, in artificial light. The sensitivity of the S5600 to filtered UV light alone is insufficient for KAP at reasonable shutter speeds.

The Hoya U-360 filter lets through much less IR but appears not to be readily available in larger sizes.  The ideal 2 inch Baader U-filter at $265+ is a little expensive, but worth it, especially if you have dedicated UV optics, and a specific application. Cheaper combinations of a Schott BG38 with either a U-330 or Tiffen 18A have also been suggested or U-330 with a BG-40 (or XNiteBP1).

High shutter speed UV KAP would probably require a dedicated digital camera with UV optics and a higher UV sensitivity than currently available. The limited continuous shooting mode of the S5600 means that we are unlikely to use it for KAP.  However, an unmodified Fuji F30fd has proved useful in low shutter speed PAP and KAP - see June 2009 update below.

The modified Optio E-35 and Fuji S5600 cameras will be useful aids in our work deciphering old manuscripts using either IR or UV.

We would like to thank Mila for enabling us to take up low-cost KAP in the near infra-red.

Modified Fuji S5600 and B+W 403 IR/UV pass filter.

ISO 200, f3.2, 1/60 sec (NB IR foliage and UV sky)

An uninspiring first attempt at IR (dominant)+UV KAP.


 June 2009

First attempt at UV pole aerial photography (PAP)

This PAP photo was taken with an unmodified Fuji F30 camera using the continuous shooting mode. However, at this shutter speed (1/4 sec), most images were blurred. This is not a problem with a digital camera when you are capturing hundreds of images in a short period of time.

Light seepage is visible at the top corners because of the simple way the filters were mounted. This has been corrected.

The pole was mounted on a ground stake. Motion blur is minimised if the shutter fires when the camera is at, or near, the maximum amplitude of swing.  Conditions for this are optimal if the shutter speed is fast (and the rate of taking photos is high) compared with the period of oscillation of the camera.  This is also true for lateral movement of a camera suspended from a kite line.  A rigid pole would be perfect, as would most poles in wind-free conditions, but a near rigid pole would have a higher frequency of oscillation than a less rigid one.

Our pole photos suggest that if a higher shutter speed is used with the Fuji F30 camera, similar UV photos should be possible using a kite, albeit with a low success rate.

Unmodified Fuji F30 with Hoya U-360 UV-pass+Schott BG39 IR-block, 25mm filters, ISO 800, f2.8, 1/4 sec.


July 2009

Fuji F30 with Hoya U-360 UV-pass + Schott BG39 IR-block, 25mm filters, ISO 1600, f2.8, 1/20 sec  KAP July 2009

First attempt at UV kite aerial photography (KAP)

The image to the left is one of the best of 600 images we obtained in our first ~20 minute flight with the UV filter assembly.  The wind was gusting, and the sun was often obscured by clouds, giving a 'success' rate (at this level of quality) of about 1%. The shutter speed was set independently of the photometry, which indicated 1/4sec and f2.8 at ISO1600. 

Taking kite aerial photos of manmade objects (buildings, roads etc) or rocks is much easier than flora because of the higher levels of reflected UV light.  Differential imaging of foliage is more problematic, but not impossible.

We are looking at the use of other filter combinations to maximise the shutter speed. Then we will consider other available compact cameras. For investigative archaeological KAP, the problems of a high ISO (with associated image noise) are less important than achieving a higher shutter speed. Fortunately, cameras are continuously improving, with lower noise/higher sensitivity, image stabilization and wider apertures.

Cameras with simple lenses are much better for UV work (including phones and PVRs).

The cheapest approach may be to use a focus-free, pinhole digital camera (1) (2) (3) (Lens cap pinhole) with the hot mirror removed. This we have not tried but it would probably require high levels of UV light. The advantage of the absence of a UV attenuating lens is reduced by the small effective aperture of the pinhole.

In our set-up to the right, the 25mm filters are joined around their edge with a strip of aluminium adhesive tape. Adhesive tack was used to attach the filter assembly to the front of the lens mount, taking care not to impede lens movement.

For suitable, unmodified cameras, this low shutter speed approach can be applied to infra-red photography (using an R72 filter) where images will be much clearer than in the UV.  Also, sensitivity to IR is usually higher than to UV.

If you have a professional interest in digital UV KAP the publications of Geert Verhoeven are recommended.

Unmodified Fuji F30fd ultraviolet set-up, with Hoya U-360 UV-pass+Schott BG39 IR-block, 25mm filters.


August 2012

Near ultraviolet KAP (video still)

Etna Brickworks, Armadale.

Looking westwards towards the levelled, former site of the pan mills (centre top).

The success rate is vastly improved by making videos in the near UV and then selecting the best stills, as with thermal imaging. Although the resolution of images obtained this way is lower than with normal images the results are more than adequate for archaeological needs.