Abstract: Forensic latent fingerprint photography, performed with evidence that has a substrate that is not uniform in terms of shape, color, and so forth, requires the use of various techniques. This report will introduce the multiple exposure method using digital photography and computerized image processing using layers methodology.

Background

Latent fingerprint (latent) photography normally requires the use of various lighting techniques to overcome the problems encountered with some surfaces upon which latents are discovered. Sometimes it is difficult (or impossible) to capture, in a single exposure, all of the detail using a single lighting technique [1]. This problem is frequently encountered on curved surfaces because multiple light angles cannot provide the optimum simultaneous illumination to all portions of the print. Problems may also be encountered when multiple excitation wavelengths or different filters are needed to provide proper exposure for different areas in the same latent.

In conventional forensic photography, there are various solutions to the problems mentioned above. For example, when a light source does not cover the total area of the latent, the painting with light technique can be used [2]. Another technique that is used in such cases is superimposition (i.e., double exposure of two images, one on top of the other, exposed on the same film frame). This technique is complex and is not frequently used in forensic photography. On the other hand, the digital multiple exposure technique is simple and productive because the merging of multiple images is more easily controlled in the computer. To the best of our knowledge, this digital method is a new application of an old technique and has not been previously reported.

Introduction of the Method

The main advantage of multiple images combined into one image is that this allows improved visualization of selected portions of a latent without affecting the rest of the image. The layers are similar to transparency pages that can be superimposed or placed one on top of the other. When there is an image on an upper layer, it is possible to see the lower layer through the use of a layer mask.

The multiple exposure technique is the superimposition of two or more separate photographs of the same latent, taken with a digital camera, while the physical position of the camera and the subject (the latent) are absolutely steady.

Because several parameters may vary, the entire process must be fully documented. The parameters that may be changed between the different photographs include the light source angle, the light source, the wavelength of illumination, and the exposure. Ideal parameters vary according to the topography of the surface area, the nature of the surface, or the location of the latent within that area.

Because the photographs are taken with a digital camera, it is very important to keep the file format (the RAW format should be used when available) and size constant. The reason for this is that a change in file size may prevent the precise alignment between the photographs when they are superimposed on separate layers.

After the photographs are taken and a separate original file is retained for each photograph, a composite Photoshop Document (PSD) is created with the Adobe Photoshop software; this file format will allow a separate layer for each image [3]. (Other image processing programs may have similar functions and capabilities.) When creating the composite file, it is important to understand that the alignment of the different layers is critical.

The next stage is to create masks for each layer. The masks for each layer can be created to either hide the selected area or to reveal only the selected area. The image areas can be selected with any of the conventional tools (e.g., the Marquee, Lasso, Magic Wand, or Paintbrush). On each layer, the layer must be masked to allow the latent area that was optimally captured in the image to be displayed, hiding the other areas that were captured but shown on different layers. In cases where the information in a certain area within two or more layers is complementary, there is an option to specify the top layer mask as a percentage (e.g., 50%), so the result will include as much information from each layer as possible.

Often the final results can be improved by changing the order of the layers, thus creating an improved blending of the composite image. Other options allow for the merging of layers, the addition of adjustment layers, and the use of layer styles.

The last phase of the process includes achieving uniformity in all of the layers by adjusting colors, brightness, and contrast; merging all the layers; and converting the final file to grayscale. The result is an image of the latent that is a computerized merge of the best portion of each image taken in order to display the maximum information available within the original latent fingerprint.

Different Exposures and Photo Superimposition Technique

This technique is recommended in cases where the object includes strong differences in the brightness levels of various areas within the latent, and the exposure differences are three aperture stops or more. Usually this phenomenon occurs when the latent appears partially upon a dark area and partially upon a light area. The quality of detail in the dark areas should be checked before using this technique.

Case Study

In this example, a latent print was developed on a beer can in a murder investigation, using basic yellow 40 (Figure 1). The latent fingerprint was located on top of the can, on a background that was part dark and part light, so the difference between the correct exposures was 3 to 4 aperture stops.

Figure 1
Beer can evidence.

In the classic method, we used a conventional camera, centerweighed light meter, Plus-X 125 ASA black and white film, f/11-16 aperture using the shutter aperture priority, 450 nm wave-length Polilight illumination (which gives fluorescence), and an orange filter. The result was insufficient because of the difference in f luorescence that leads to the absence of information in the dark surface area (Figure 2).

To solve this problem, we took two photographs using a Fujifilm S1 digital camera and Micro-Nikkor AF lens. The two exposures were taken while the position of the camera and the position of the beer can were not changed. The first photograph was taken with the correct exposure (according to the fluorescence of the bright area of the can surface), and the second photograph was overexposed by 3 stops to record the detail in the dark area.

Using Photoshop, we opened a PSD file. The first photograph (normal exposure) was defined as the background layer and the second photograph, which recorded the detail in the dark area, was pasted on top of the background as a second layer. Working on the background layer, the letters on the beer can were selected using the Magic Wand tool. (The dark uniform coloration of the letters made them easy to select with this tool.) After a selected area is defined, the selection borders appear on any active and visible layer and can be used on the second layer. (As explained above, this layer contained the latent information on top of the dark letters because this photograph was overexposed).

To get rid of the unusable information on the second layer, we used the Add Layer Mask and Reveal Selection commands and hid the overexposed portion of that image (the areas other than the selected letters). By using the Levels command, the two layers were brought to the same level of brightness and contrast. At this stage, the two layers were ready for merging into one layer, containing the full information of the latent fingerprint (Figure 3).

Figure 2
Fingerprint luminescence at normal exposure.

Figure 3
Composite of two different exposures.

Merging Photographs that were taken from Different Light Source Angles

This technique is recommended in cases where the latent fingerprint appears on top of objects that do not have a uniform surface topography (e.g., a convex or concave surface). If the surface is shiny, then the ref lection problem can be corrected by controlling the angle of a single light. On this kind of surface, it is usually impossible to see the complete latent at the same time, and the light source must be adjusted in order to emphasize different areas of the print. Sometimes a different light source angle is not sufficient and different light sources must be used. The same technique procedures would apply.

Case Study

This example is from a terrorist attack investigation. Prior to fingerprint processing, a visible fingerprint was located on the aluminum clip on a cellular phone holster. Because the print was located inside the concave portion of the clip (Figure 4), it was impossible to see the whole of it from a single angle, and only parts of the information could be made visible by changing the angle of illumination.

Figure 4
Position of fingerprint on cellular phone holster.

In addition to this difficulty, and to avoid any risk that the print could be damaged and lost during the chemical development process, it was photographed prior to any processing techniques. All photographs were taken while the location of the camera and the subject remained unchanged, and only the angle of the light source was adjusted. We used a Fujifilm S1 Pro, Micro Nikkor AF 60 mm lens (aperture f/16), and concentrated white light (using a Polilight device).

The resulting photographs provided views of the two different parts of the latent fingerprint on the clip (Figures 5 and 6). The purpose of the computerized processing was to merge the two parts of the latent into one complete latent fingerprint with no distortions.

Figure 6 was selected as the background layer and Figure 5 as the top layer. A mask was created for the top layer using the commands Add Layer Mask and Hide All. After creating the hidden mask, we selected a brush with the needed characteristics of size and softness. It was necessary to choose white as the front color, in order to use the brush efficiently to “scratch” the opaque layer, which exposes the information from the hidden layer underneath. (By selecting the front color as black, the mask can be repainted to restore the masking effect.)

These steps require a lot of care; therefore, it is recommended to work with a small diameter brush. Because the two parts of the latent were overlapping (Figures 5 and 6), we achieved a smooth joining of the fingerprint ridges. The Levels command was used to reach the same levels of brightness and contrast in the two layers. The process was completed by merging the two layers and saving the final result as a flattened file (Figure 7).

Figure 5
First part of fingerprint.

Figure 6
Position of fingerprint on cellular phone holster.

Figure 7
Composite of two light source angles.

Conclusion

Implementing the multiple exposure techniques in digital photography and using image processing, as described in this paper, provides a lot of opportunities to produce better photographs of latent fingerprints. We strongly recommend using this technique when taking photographs of latent fingerprints on ref lecting surfaces (e.g., metals or glass) or fluorescent surfaces, especially when the use of one light source or light source angle is not sufficient for obtaining all the details from the latent.

Great care should be observed in maintaining the chain of evidence. For every job, we recommend the use of a separate source folder that should contain the original untouched photographs. All digital photographs should be processed using copies of the original photographs, preferably in a separate folder named “image processing”. The final processed image should be inspected and compared to the source images to ensure that no artifacts were introduced. The person doing the processing should be able to show in court the origin of every part of a composite image and explain the techniques involved.

Acknowledgment

The authors wish to thank Ms. Jill Segal for her comments and assitance in editing and preparing this report.

For further information, please contact:


Alan Chaikovsky or Uri Argaman
Scientific Officer, Photography Laboratory
Division of Identification and Forensic Science (DIFS)
Israel Police, National H.Q.
Jerusalem 91906
alan8@bezeqint.net

References

1. Langford, M. The Basics of Photography; Focal Press: London, 1982.
2. Tal, S.; Ilan, M. Photography Lexicon; Focus Press: Tel-Aviv, 1990.
3. Adobe Photoshop 7 User Guide; Adobe Systems Inc.: San Jose, 2002.
4. Russ, J. C. Forensic Uses of Digital Imaging; CRC Press: Boca Raton, FL, 2001.