J Forensic Sci, March 2015, Vol. 60, No. 2 doi: 10.1111/1556-4029.12686 Available online at: onlinelibrary.wiley.com


Glenn Porter,1 Ph.D.; Robert Ebeyan,1 B.Sc. (Hons.); Charles Crumlish,2 B.A. (Hons.); and Adrian Renshaw,2 Ph.D.

A Novel Method for the Photographic Recovery of Fingermark Impressions from Ammunition Cases Using Digital Imaging

ABSTRACT: The photographic preservation of fingermark impression evidence found on ammunition cases remains problematic due to the

cylindrical shape of the deposition substrate preventing complete capture of the impression in a single image. A novel method was developed for the photographic recovery of fingermarks from curved surfaces using digital imaging. The process involves the digital construction of a complete impression image made from several different images captured from multiple camera perspectives. Fingermark impressions deposited onto 9-mm and 0.22-caliber brass cartridge cases and a plastic 12-gauge shotgun shell were tested using various image parameters, including digital stitching method, number of images per 360° rotation of shell, image cropping, and overlap. The results suggest that this method may be successfully used to recover fingermark impression evidence from the surfaces of ammunition cases or other similar cylindrical surfaces.

KEYWORDS: forensic science, forensic photography, fingerprint, ammunition cases, digital imaging, image stitching

Fingermarks detected on physical evidence and at crime scenes are considered within forensic investigation as a valuable form of identification evidence. A primary fingerprint preservation method used in forensic science is photographic documentation; however, this process remains problematic for fingermarks found on ammunition cases due to their cylindrical shape (1–4). The curved surface of ammunition cases presents a number of challenges. The principal difficulties involve issues with lighting curved reflective surfaces and the need to record fingermarks in their entirety, including any sections of the fingermark that may be wrapped around the casing and hidden from the view of the camera (5). A literature survey revealed only a limited number of solutions have been published that address the issue of photographing fingermarks from substantially curved surface including Pippin’s “Rotorgraph” (5) and Williams & McMurray’s scanning Kelvin probe (6). This paper examines the development of a novel method for the recovery of fingermark impression evidence from the curved surface of ammunition cases. The method applies digital imaging to combine multiple images captured from a number of different camera perspectives and presenting the entire fingerprint impression result in a single image.

1 College of Arts, Society & Education, James Cook University, PO Box 6811, Cairns, Qld 4870, Australia. 2 School of Science & Health, University of Western Sydney, Locked Bag 1797, Penrith, NSW 2751, Australia. Received 6 May 2013; and in revised form 28 Mar. 2014; accepted 15 April 2014.


Materials and Methods Postfired brass 0.22 and 9-mm caliber cartridge cases and plastic nonribbed 12-gauge shotgun shells were tested during this research. The variation in these specimens provided differences in size and degree of curvature. Prior to fingerprint deposition, all samples were cleaned to remove any pre-existing fingermark residue or surface contaminants by systematically washing the samples in a series of solutions described by Bond (7). Samples were first washed in a dilute solution of commercial detergent, washed in a solution of distilled H2O, followed by a solution of acetone, and finally, another solution of distilled H2O. Fingermark impressions were deposited by a male donor lightly pressing a finger onto the sample surface and then rolled through to the fingertip resulting in the deposition of latent fingermarks that wrapped around the sample surface. The latent impressions were developed using fluorescent fingerprint powder (Optimum Technology Lightning Blitz Red Fluoro-Magnetic Print Powder and Sirchie REDescentTM No. LL701 Latent Fingerprint Powder), and adhesive fluorescent linear scales were applied to the samples during the photography procedures. Linear scales were applied directly to the 12-gauge shotgun shell sample around the brass head, below the developed fingermark. The linear scales were added to the specimens to examine how successfully the combined images merged while maintaining consistent spacial elements within the results. Incorrect spacial elements would be detected as inconsistencies between the linear scale divisions within the image. A ManfrottoTM panoramic tripod head (QTVR) was used as the specimen mounting stage and rotation device. The traditional function of the panorama tripod head is to enable a mounted camera to revolve on its optical center assisting in the construction of © 2014 American Academy of Forensic Sciences


panoramic images. In this method, the panoramic tripod head is used as a rotatable specimen platform, enabling 360° photographic capture of the ammunition cases, while the camera remained stationary. The device is equipped with several degrees of rotation settings that allowed for the selection of a determined number of images per 360° revolution. Table 1 provides details of the QTVR tripod head settings. Photoluminescence method (2) was used for photographing the ammunition case specimens to increase fingermark contrast and assist with eliminating lighting artefacts caused by the reflective surface of the substrate (metal). A Canon EOS digital SLR camera equipped with a 100-mm macro lens with a RofinTM OG590 barrier filter attached was used for the original capture. A monochromatic light source was used to illuminate the samples and excite the fluorescent fingerprint powder (using appropriate central bandwidth depending on the fluorescent powder used). The PolilightTM light guide was positioned next to the camera lens to provide near coaxial illumination so that the sample surfaces were evenly lit (2) preventing uneven lighting and light fall-off across the specimen surface. The photoluminescent fingermark must also produce a consistent level of brightness across the entire specimen when photographed for successful combination of the individual images. The OG590 barrier filter increased the contrast of the specimen by transmitting the fluorescence of the fingermark while absorbing any reflected light from the background. Figure 1 provides a diagram of the photographic setup. A number of image parameters were investigated regarding the successful stitching of several images to form a single fingermark impression image, including:

• • • • •

The number of images per 360° rotation, Percentage of image overlap between each image, Percentage of cropping, Image enhancement, and Stitching software parameters.

After each photograph was captured, the panoramic tripod head was rotated at the desired rotation setting until an entire revolution of the specimen was completed and photographed at several different viewpoints relative to the camera position. Each photograph was taken perpendicular to the specimen and with a degree of overlap required for successful image stitching (Fig. 2). Digital image processing and stitching was performed using Adobe PhotoshopTM. Each image underwent contrast enhancement using Photoshop’s “curves” function followed by sharpening. Images were then cropped, which involved cropping the




ammunition cases to form a narrow central band that spanned 40% of the casing’s width. This resulted in the images taking on the appearance of narrow strips (see Fig. 3). These cropped images were then used as the source for image stitching. Cropping the images in this manner removed the effects of perspective distortion, slight changes in impression size, and image compression, caused by the curvature of the ammunition cases and the oblique viewpoint toward the edge of the casing. Cropping the image also negated any sharpness issues caused by insufficient depth of field experienced when performing close-up photography. The cropping process also maintained a degree of overlap with consecutive images (left and right). The PhotomergeTM function available in Photoshop is traditionally used for constructing panoramic images through the combination or stitching of several images that contain overlapping regions to form a single composite image (8). In this method, PhotomergeTM function was used to combine the images to form a single image that represented the entire fingermark present on

FIG. 1––Photography equipment set up. A = cartridge case with treated fingermark, B = sample holder with applied linear scale (inverted blank cartridge case), C = brass spigot used as sample stage, D = ManfrottoTM panoramic tripod head, E = lens barrier filter (RofinTM OG590), F = 100 mm macro lens, G = Canon EOS 50DTM digital SLR camera, H = monochromatic light source (RofinTM Polilightâ PL500, I = light guide holder, J = camera tripod.

TABLE 1––Image parameters; Setting available on the ManfrottoTM panoramic tripod head that enable the associated number of images to be captured of the subject per revolution. Degree of rotation of the tripod head each time it was manually rotated. No. Images/360° 4 6 8 10 12 15 18 24 36

Tripod Head Setting

Degree of Rotation (°)

n4 n6 n8 n10 n12 n15 n18 n24 n36

90 60 45 36 30 24 20 15 10

FIG. 2––Cropped images displaying overlapping before stitching combines images to form a single image.



the cylindrical surface of the ammunition cases. When combining images, Photomerge’s “reposition only” option was selected, as this option permitted images to be stitched together without compromising image integrity by stretching or skewing component images during the stitching process. The “blend images together” option was also selected as this option produced results that masked any subtle differences in exposure between each of the individual images. Results and Discussion All digitally stitched images were compared with inked exemplar fingermark impressions using a known source as the ground truth. The photographic recovery of fingermark impressions were determined to be successful if a set of images obtained for a particular image parameter (e.g., percentage of overlap, degree of cropping, and number of images per rotation) combined producing a result that excluded the presence of any image distortion or artefacts and depicted what appeared as a fair representation of the fingermark. Fingermark impression stitching was deemed to be unsuccessful if the set of images failed to combine or if distorting artefacts were observed as a result of their merging.

Artefacts include missing or repeated ridges when the stitching processing was applied. The results indicated that fingermark impressions may be successfully recovered from all three types of the ammunition case samples examined. Fingermark impressions were successfully recovered from both the 9-mm caliber cartridge case and the 12gauge shotgun shell using the 15, 18, 24, and 36 images per 360° rotation. Fingermark impressions were successfully recovered from the 0.22-caliber cartridge case using the 12, 15, 18, 24, and 36 images per 360° rotation (see Table 2). Table 3 lists the results and image parameters of the experimental samples, and Fig. 4 is an example of a final result. There are a limited number of previous solutions for recovering fingermarks wrapped around ammunition casings expressed in the literature (5,6), and some methods also offer other advantages for the recovery of this form of physical evidence. The advantage of this fingermark recovery method is that it is achievable using readily available photographic equipment, and the method can capture high-resolution images of the fingermark TABLE 3––Summary of the image parameters used in the stitching method and the best results achieved. Image Parameter



i. Min number of images per 360° rotation

0.22 9 mm 12-gauge 0.22 9 mm 12-gauge 0.22 9 mm 12-gauge

12 images/360° 15 images/360° 15 images/360° 35 percent 57 percent 48 percent 40 percent 40 percent 40 percent Photoluminescence mode Fluoro-Magnetic print powder Rofin OG590 barrier filter Monochrome lighting at 415 nm central wavelength Adobe Photoshop Photomerge application “reposition only” and “blend images together” settings used

ii. Percentage of image overlap between each image iii. Percentage of cropping iv. Image enhancement

v. Stitching software

FIG. 3––Cropping specimen in preparation of image stitching.

TABLE 2––This table illustrates the number of rotations for fingermarks that successfully stitched (U) without any obvious image artefacts. The numbers represent the total number of images taken per 360° rotation. Successful Stitching Results No. Images/360° Sample 9 mm 0.22 caliber 12-gauge

4 ✗ ✗ ✗

6 ✗ ✗ ✗

8 ✗ ✗ ✗

10 ✗ ✗ ✗

12 ✗ U ✗

15 U U U

18 U U U

24 U U U

36 U U U

FIG. 4––Final result constructed (stitched) image from multiple images.


in a relatively short amount of time. This novel method achieved similar results to that offered by Pippin (5) using contemporary digital imaging approach to the same problem. Like Pippin’s results, this method may also be applied to other cylindrical items that contain fingermarks like screwdrivers, pens, door handles, light globes, electrical fuses, and others. Conclusion A novel method for the photographic recovery of fingermark impression evidence from the surfaces of ammunition cases using digital imaging was developed. The method was used to effectively recover fingermarks from the curved surfaces of 12gauge, 9-mm and 0.22-caliber ammunition cases. The results suggest that the method is capable of successfully recovering fingermark evidence from the cylindrical surfaces of ammunition cases in a single image. This method does, however, produce a constructed image made from several individual images. This conceptual reference to the legitimacy of an image that is constructed from several images may raise concerns in relation to concepts of photographic truth and reliability (9,10). Further research is imperative for this method to determine whether the constructed image parameters detailed in this paper can reproduce faithfully the complex fingermark pattern made from the source for the purpose of forensic examination.




References 1. Porter G. A new theoretical framework regarding the application and reliability of photographic evidence. Int J Evid Proof 2011;15:26–61. 2. Porter G. Photography: marks, impressions and documents. In: Jamieson A, Moenssens A, editors. Wiley encyclopedia of forensic science. New York, NY: Wiley & Sons, 2009;2036–57. 3. Hawthorne MR. Fingerprints: analysis and understanding. Boca Raton, FL: CRC Press, 2008. 4. Saferstein R. Criminalistics: an introduction to forensic science, 9th edn. Upper Saddle River, NJ: Pearson Prentice Hall, 2007. 5. Pippin T. The rotorgraph. J Fam Issues 1995;45(6):612–7. 6. Williams G, McMurray N. Latent fingermark visualisation using a scanning kelvin probe. Forensic Sci Int 2007;167(2–3):102–9. 7. Bond JW. Visualization of latent fingerprint corrosion of metallic surfaces. J Forensic Sci 2008;53(4):812–22. 8. Evening M, Schewe J. Adobe Photoshop CS5 for photographers: the ultimate workshop. London, U.K.: Focal Press, 2011. 9. Porter G. Visual culture in forensic science. Aust J Forensic Sci 2007;39 (2):81–91. 10. Porter G, Kennedy M. Photographic truth and evidence. Aust J Forensic Sci 2012;44(2):183–92. Additional information and reprint requests: Glenn Porter, Ph.D. Associate Professor in Photomedia College of Arts, Society & Education James Cook University PO Box 6811, Cairns, Qld 4870 Australia E-mail: [email protected]

A novel method for the photographic recovery of fingermark impressions from ammunition cases using digital imaging.

The photographic preservation of fingermark impression evidence found on ammunition cases remains problematic due to the cylindrical shape of the depo...
595KB Sizes 0 Downloads 4 Views