Aspects of 3D surface scanner performance for post-mortem skin documentation in forensic medicine using rigid benchmark objects (original) (raw)

Abstract

Background: Patterned light 3D scanning has historically been targeted towards industrial and manufacturing applications. Forensic 3D skin surface scanning is relatively new and appears to contain aspects of off label usage. Based on how patterned light scanning has been published to work, we assumed that naturally rough surfaces' 3D scan validity to improve with extensive calibration of such a 3D-scanner whereas we assumed the same not to be true for industrially smooth surfaces. Using rigid benchmark objects matching aspects of typical post-mortem skin injuries and an object with smooth plastics surface, that hypothesis was tested.

Figures (9)

Fig. 1. Benchmark objects used: (a) photo of a sheep's skull containing small intrinsic bone surface features and tool marks, including add-ons such as boreholes, countersink drillings, red and black felt pen marks. (b) Photo of a sandstone conglomerate featuring different inclusions, some of dark light absorbing quality, some highly reflective. (c) Remote control as example of an industrial surface; in addition, it was coated with white anti-reflective spray (partly wiped off after 3D scan and before photography). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) We employed a QTSculptor PT-M1280 surface scanner by Polygon Technology (Darmstadt, Germany). The scanner projects collimated We investigated a 3D scanner's performance for what is properly called off label application domain of 3D skin surface documentation in forensic pathology. Initial testing revealed a questionable performance regarding fidelity of rough surface representation while industrially smooth objects appeared to get

Fig. 1. Benchmark objects used: (a) photo of a sheep's skull containing small intrinsic bone surface features and tool marks, including add-ons such as boreholes, countersink drillings, red and black felt pen marks. (b) Photo of a sandstone conglomerate featuring different inclusions, some of dark light absorbing quality, some highly reflective. (c) Remote control as example of an industrial surface; in addition, it was coated with white anti-reflective spray (partly wiped off after 3D scan and before photography). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) We employed a QTSculptor PT-M1280 surface scanner by Polygon Technology (Darmstadt, Germany). The scanner projects collimated We investigated a 3D scanner's performance for what is properly called off label application domain of 3D skin surface documentation in forensic pathology. Initial testing revealed a questionable performance regarding fidelity of rough surface representation while industrially smooth objects appeared to get

Fig. 2. Real skin sample (pig skin from an animal killed for nutritional purposes) illustrating the non-suitability of realistic skin samples for benchmarking 3D scanner resolution for rough surfaces. Photographs showing decay 3 days (a) and 6 days (b) post-mortem. 3D surface scans at 3 days (c) and 6 days (d) (bar 1 cm) post-mortem provide surfaces of similar appearance, but even at good overlap positioning a distance map (e) between the two digitized surfaces shows that they are not congruent with patches of divergence exceeding 2 mm.

Fig. 2. Real skin sample (pig skin from an animal killed for nutritional purposes) illustrating the non-suitability of realistic skin samples for benchmarking 3D scanner resolution for rough surfaces. Photographs showing decay 3 days (a) and 6 days (b) post-mortem. 3D surface scans at 3 days (c) and 6 days (d) (bar 1 cm) post-mortem provide surfaces of similar appearance, but even at good overlap positioning a distance map (e) between the two digitized surfaces shows that they are not congruent with patches of divergence exceeding 2 mm.

Fig. 3. Side-by-side comparison of real forensic skin pathology photographs (a, b, d, e, g, h) and benchmark object feature photograph (c) or 3D scan (f, i): deep facial abrasion after sliding over a rough surface (a, b) containing various highly reflective surface regions, patchy dark discoloration as well as bumpy appearance. These surface shape elements are represented on a similar scale on the rock surface that we used as benchmark object (c). Superficial abrasions as found in the vicinity of gunshot entry wounds (d, e: arrow) are also present on skull surface (f: arrow) used as benchmark object. Curved (g: arrow) and straight (h: arrow) wound edges as found in stabs from a serrated (g) or straight (h) knife blade are represented by a bony suture of the skull (i: arrow) used as benchmark object (bar 1 cm).

Fig. 3. Side-by-side comparison of real forensic skin pathology photographs (a, b, d, e, g, h) and benchmark object feature photograph (c) or 3D scan (f, i): deep facial abrasion after sliding over a rough surface (a, b) containing various highly reflective surface regions, patchy dark discoloration as well as bumpy appearance. These surface shape elements are represented on a similar scale on the rock surface that we used as benchmark object (c). Superficial abrasions as found in the vicinity of gunshot entry wounds (d, e: arrow) are also present on skull surface (f: arrow) used as benchmark object. Curved (g: arrow) and straight (h: arrow) wound edges as found in stabs from a serrated (g) or straight (h) knife blade are represented by a bony suture of the skull (i: arrow) used as benchmark object (bar 1 cm).

Fig. 4. Sheep skull benchmark object. (a) Photo. (b) Matching view of 3D model. Details of the scratch contained on the surface were reproduced photographically (c) and using 3D scanner (d). See text for feature description. Bar is 5 mm. Qualitatively (Figure references in brackets), similarity of scan- ner generated data can be appreciated by comparing photographs (4a and c, 5a and c, 6(photo)) with matching 3D scan data (4b and d, 5b and d, 6(EC)). While the overall appearance of the scanned 3D data seemed to match the photographed object at first glance, individual surface patches need to be compared as errors manifest themselves on a patch-by-patch level. A bone surface scratch (4a(1) and c(1)) was well discernible on 3D scan derived data (4b(1) and d(1)). The roughness of the finely granular bone surface (4a(3), 6a(photo) and b(photo)) appeared to be adequately repre- sented on 3D scan generated data (4b(3), 6a(x) and b(EC)). Small grooves present on the sandstone rock (5a(1) and c(3): between arrows) were discernible on 3D scan derived data (5b(1) and d(3): between arrows); two distinct humps present on the rock (5a(2): arrows) could be differentiated on 3D scan data (5b(2)). Borehole rims (4a(2)) were available for inspection in 3D scan data (4b(2)). Shiny quartz inclusions (5a(4) and c(5)) were still represented by 3D scan data (5b(4) and d(5)). Overall the 3D scanner generated surfaces contained rich details, good representation of highly We obtained minimally calibrated (MC) scans as well as extensively calibrated (EC) 3D data of skull and remote control containing comparable polygon counts (Table 2, top). Scan times were below 15 min for each object.

Fig. 4. Sheep skull benchmark object. (a) Photo. (b) Matching view of 3D model. Details of the scratch contained on the surface were reproduced photographically (c) and using 3D scanner (d). See text for feature description. Bar is 5 mm. Qualitatively (Figure references in brackets), similarity of scan- ner generated data can be appreciated by comparing photographs (4a and c, 5a and c, 6(photo)) with matching 3D scan data (4b and d, 5b and d, 6(EC)). While the overall appearance of the scanned 3D data seemed to match the photographed object at first glance, individual surface patches need to be compared as errors manifest themselves on a patch-by-patch level. A bone surface scratch (4a(1) and c(1)) was well discernible on 3D scan derived data (4b(1) and d(1)). The roughness of the finely granular bone surface (4a(3), 6a(photo) and b(photo)) appeared to be adequately repre- sented on 3D scan generated data (4b(3), 6a(x) and b(EC)). Small grooves present on the sandstone rock (5a(1) and c(3): between arrows) were discernible on 3D scan derived data (5b(1) and d(3): between arrows); two distinct humps present on the rock (5a(2): arrows) could be differentiated on 3D scan data (5b(2)). Borehole rims (4a(2)) were available for inspection in 3D scan data (4b(2)). Shiny quartz inclusions (5a(4) and c(5)) were still represented by 3D scan data (5b(4) and d(5)). Overall the 3D scanner generated surfaces contained rich details, good representation of highly We obtained minimally calibrated (MC) scans as well as extensively calibrated (EC) 3D data of skull and remote control containing comparable polygon counts (Table 2, top). Scan times were below 15 min for each object.

Fig. 5. Rock benchmark object. (a) Photo. (b) Matching view of 3D model obtained using QTSculptor scanner. A detailed region of this surface was reproduced photographically (c) and using QTSculptor (d). See text for feature description. Bar is 5 mm.

Fig. 5. Rock benchmark object. (a) Photo. (b) Matching view of 3D model obtained using QTSculptor scanner. A detailed region of this surface was reproduced photographically (c) and using QTSculptor (d). See text for feature description. Bar is 5 mm.

Fig. 6. Experimental subjective evaluation was conducted using a set of 20 comparisons just as the four visual objects (a-d) displayed here, evaluated by 13 participants. Two calibration types (EC, extensively calibrated; MC, minimally calibrated) were used. For each of the rows, 3D model EC and MC were presented in random sequence to participants wh« had to select either a preference (‘x’ or ‘y’) or ambivalence (‘x’ and ‘y’ equally good) as a forced choice. In this illustration, EC labels the extensively calibrated 3D scan, MC labels thi minimally calibrated 3D scan for all four visual objects (a—d). (a) Bone surface structure featuring finely granular roughness, and superficial indents, some of which exhibit subtle edge (arrows). (b) Bone surface structure featuring finely granular roughness and a curved bone layer edge (arrow). (c and d) Industrial surface exhibits curved and straight edges, a smal protrusion (arrow) and mostly flat surfaces (uneven spray coating left some particles that can be discerned on both 3D scans). Bar is 5 mm.

Fig. 6. Experimental subjective evaluation was conducted using a set of 20 comparisons just as the four visual objects (a-d) displayed here, evaluated by 13 participants. Two calibration types (EC, extensively calibrated; MC, minimally calibrated) were used. For each of the rows, 3D model EC and MC were presented in random sequence to participants wh« had to select either a preference (‘x’ or ‘y’) or ambivalence (‘x’ and ‘y’ equally good) as a forced choice. In this illustration, EC labels the extensively calibrated 3D scan, MC labels thi minimally calibrated 3D scan for all four visual objects (a—d). (a) Bone surface structure featuring finely granular roughness, and superficial indents, some of which exhibit subtle edge (arrows). (b) Bone surface structure featuring finely granular roughness and a curved bone layer edge (arrow). (c and d) Industrial surface exhibits curved and straight edges, a smal protrusion (arrow) and mostly flat surfaces (uneven spray coating left some particles that can be discerned on both 3D scans). Bar is 5 mm.

Results of scanner performance and resolution evaluation. Table 1

Results of scanner performance and resolution evaluation. Table 1

Impact of extent of calibration on perceived surface quality. Subjective blinded experimental evaluation of 2 x 10 items (10: skull; 10: remote control) by 13 individuals yields a statistically significant preference for extensively calibrated scans of rough surfaces, whereas such a predominance was not found for industrially smooth surfaces (comparing preference count distribution of + rough surfaces against preference count distribution of + industrially smooth surfaces using G2 Likelihood Ratio Chi-square statistic with p < 0.0001). EC, extensive calibration; MC: minimal calibration.

Impact of extent of calibration on perceived surface quality. Subjective blinded experimental evaluation of 2 x 10 items (10: skull; 10: remote control) by 13 individuals yields a statistically significant preference for extensively calibrated scans of rough surfaces, whereas such a predominance was not found for industrially smooth surfaces (comparing preference count distribution of + rough surfaces against preference count distribution of + industrially smooth surfaces using G2 Likelihood Ratio Chi-square statistic with p < 0.0001). EC, extensive calibration; MC: minimal calibration.

Fig. 7. Logistic regression model fit for size of single shape feature (x-axis) against recognition rate for 220 items containing patches of naturally rough surfaces (skull, rock).

Fig. 7. Logistic regression model fit for size of single shape feature (x-axis) against recognition rate for 220 items containing patches of naturally rough surfaces (skull, rock).

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (61)

  1. W. Brüschweiler, M. Braun, R. Dirnhofer, M.J. Thali, Analysis of patterned injuries and injury-causing instruments with forensic 3D/CAD supported photogrammetry (FPHG): an instruction manual for the documentation pro- cess, Forensic Science International 132 (2) (2003) 130-138.
  2. J. Subke, H.D. Wehner, F. Wehner, S. Szczepaniak, Streifenlichttopometrie: a new method for the three-dimensional photorealistic forensic documentation in colour, Forensic Science International 113 (1-3) (2000) 289-295.
  3. M.I.M. Goos, I.B. Alberink, A.C.C. Ruifrok, 2D/3D image (facial) comparison using camera matching, Forensic Science International 163 (1-2) (2006) 10-17, Comparative study.
  4. M. Häusler, W. Schweitzer, W. Bär, Methodic improvement of the photographic superimposition for the identification of skulls. in: Proceedings of the XX Congress of International Academy of Legal Medicine, Budapest, Hungary, Medimond, Bologna, Italy, 2006, pp. 213-215.
  5. W. Brüschweiler, M. Braun, H. Fuchser, R. Dirnhofer, Photogrammetrische Auswertung von Haut-und Weichteilwunden sowie Knochenverletzungen zur Bestimmung des Tatwerkzeuges -grundlegende Aspekte, Rechtsmedizin 7 (3) (1997) 76-83.
  6. M.J. Thali, M. Braun, W. Bruschweiler, R. Dirnhofer, Matching tire tracks on the head using forensic photogrammetry, Forensic Science International 113 (2000) 281-287.
  7. M.J. Thali, M. Braun, J. Wirth, P. Vock, R. Dirnhofer, 3D surface and body documentation in forensic medicine: 3-D/CAD photogrammetry merged with 3D radiological scanning, Journal of Forensic Sciences 48 (2003) 1356-1365.
  8. M.J. Thali, M. Braun, W. Brueschweiler, R. Dirnhofer, 'Morphological imprint': determination of the injury-causing weapon from the wound morphology using forensic 3D/CAD-supported photogrammetry, Forensic Science Interna- tional 132 (2003) 177-181.
  9. M.J. Thali, M. Braun, R. Dirnhofer, Optical 3d surface digitizing in forensic medicine: 3d documentation of skin and bone injuries, Forensic Science International 137 (2-3) (2003) 203-208.
  10. G. Sansoni, C. Cattaneo, M. Trebeschi, D. Gibelli, D. Porta, M. Picozzi, Feasibility of contactless 3d optical measurement for the analysis of bone and soft tissue lesions: new technologies and perspectives in forensic sciences, Journal of Forensic Sciences 54 (May 6) (2009) 540-545.
  11. G. Ampanozi, T.D. Ruder, U. Preiss, K. Aschenbroich, T. Germerott, L. Filograna, et al., Virtopsy: CT and MR imaging of a fatal head injury caused by a hatchet: a case report, Legal Medicine 12 (5) (2010) 238-241.
  12. M. Grassberger, A. Gehl, K. Puschel, E. Turk, 3d reconstruction of emergency cranial computed tomography scans as a tool in clinical forensic radiology after survived blunt head trauma-report of two cases, Forensic Science International 207 (1-3) (2011) e19-e23.
  13. J. Hiss, T. Kahana, Medicolegal investigation of death in custody: a postmortem procedure for detection of blunt force injuries, American Journal of Forensic Medicine and Pathology 17 (1996) 312-314.
  14. B. Karger, J. Niemeyer, B. Brinkmann, Suicides by sharp force: typical and atypical features, International Journal of Legal Medicine 113 (2000) 259-262.
  15. J.A. Kieser, V. Bernal, J.N. Waddell, S. Raju, The uniqueness of the human anterior dentition: a geometric morphometric analysis, Journal of Forensic Sciences 52 (2007) 671-677.
  16. S. Blackwell, R. Taylor, I. Gordon, C. Ogleby, T. Tanijiri, M. Yoshino, et al., 3-D imaging and quantitative comparison of human dentitions and simulated bite marks, International Journal of Legal Medicine 121 (2007) 9-17.
  17. M.M. Baden, Homicide, suicide, and accidental death among narcotic addicts, Human Pathology 3 (1) (1972) 91-95.
  18. J. Walker, Bullet holes and chemical residues in shooting cases, Journal of Criminal Law and Criminology 31 (1940) 497-521.
  19. J. Rainio, K. Lalu, H. Ranta, A. Penttilä, Morphology of experimental assault rifle skin wounds, International Journal of Legal Medicine 117 (2003) 19-26.
  20. M. Grosse Perdekamp, S. Pollak, Elucidation of a strange gunshot injury, International Journal of Legal Medicine 119 (2005) 91-93.
  21. Durham County Superior Court, State v. Michael Peterson. CRS 024821 2001; File No. 01:1.
  22. United States District Court, United States v. O'Brian Cleary "O.C." Smith. Western Division Tennessee 2004; File No. 04-20054.
  23. H. Weber, M. Dorr, Digitization as a Means of Preservation?, European Commission on Preservation and Access, Amsterdam, 1997.
  24. W.B. Thompson, J.H. Owen, J.deSt. Germain, S.R. Stark Jr., T.C. Henderson, Feature-based reverse engineering of mechanical parts, IEEE Transactions on Robotics and Automation 15 (1) (1999) 57-66.
  25. W. Boehler, M. Bordas Vincent, A. Marbs, Investigating laser scanner accuracy, in: XIXth CIPA Symposium-Working Group 6, Antalya, Turkey, 2003, pp. 1-6.
  26. G. Schaepman-Strub, M. Schaepman, T.H. Painter, S. Dangel, J. Martonchik, Reflectance quantities in optical remote sensing-definitions and case studies, Remote Sensing of Environment 103 (2006) 27-42.
  27. J. Subke, H.D. Wehner, F. Wehner, S. Szczepaniak, Streifenlichttopometrie (SLT): a new method for the three-dimensional photorealistic forensic documentation in colour, Forensic Science International 113 (1-3) (2000) 289-295.
  28. ABW GmbH, ABW BackMapper, 2012. URL〈http://web.archive.org/web/ 20010220104843/http://www.abw-3d.de/〉.
  29. R.S. Stafford, Regulating off-label drug use-rethinking the role of the FDA, New England Journal of Medicine 358 (14) (2008) 1427-1429.
  30. F. Blais, Review of 20 years of range sensor development, Journal of Electronic Imaging 13 (1) (2004) 231-243, http://dx.doi.org/10.1117/1.1631921.
  31. Creaform, CREAFORM, 2012. URL 〈http://www.creaform3d.com/〉.
  32. TC]2, [TC]2 3D Body Scanning Technologies, 2012. URL 〈http://.tc2.com/〉.
  33. TELMAT Industrie, Symcad, 2012. URL 〈http://www.symcad.com/〉.
  34. Wicks and Wilson Limited, WWL 3D Scanning Products, 2012. URL 〈http:// www.wwl.co.uk/3dscanning-index.htm〉.
  35. GOM mbH, ATOS-Industrial 3D Scanning Technology, 2012. URL 〈http:// www.gom.com/metrology-systems/3d-scanner.html〉.
  36. Capture 3D, 3D Industrial Measurement Solutions, 2012. URL 〈http://www. capture3d.com/applications.html〉.
  37. Steinbichler, Surface Scanning, 2012. URL 〈http://www.steinbichler.de/pro dukte/surface-scanning.html〉.
  38. D. Parkot, Projekte-Zahnärztliche Werkstoffkunde und Biomaterialforschung, DigImprint, 2012. URL 〈http://tinyurl.com/9yhz8h7〉.
  39. Optonet, Optonet, 2012. URL 〈http://www.optonet.it/optoweb/ita/applicazioni. htm〉.
  40. Voxelan, Voxelan Scanners, 2012. URL 〈http://www.voxelan.co.jp〉.
  41. Geomagic, Geomagic, 2012. URL 〈http://www.geomagic.com〉.
  42. ISO, ISO 5725-5. Accuracy (Trueness and Precision) of Measurement Methods and Results-Alternative Methods for the Determination of the Precision of a Standard Measurement Method, International Organization for Standardiza- tion, Geneva, 1994.
  43. M. Petrov, A. Talapov, T. Robertson, A. Lebedev, A. Zhilyaev, L. Polonskiy, Optical 3d digitizers: bringing life to the virtual world, IEEE Computer Graphics and Applications 18 (3) (1998) 28-37, http://dx.doi.org/10.1109/ 38.674969.
  44. K. Harding, Calibration methods for 3D measurement systems, vol. Class 1, Technical Information Series, GE Research & Development Center, 2001.
  45. ISO, ISO 5725-2. Accuracy (Trueness and Precision) of Measurement Methods and Results-Basic Method for the Determination of Repeatability and Reproducibility of a Standard Measurement Method, International Organiza- tion for Standardization, Geneva, 1994.
  46. Guardbanding and the World of ISO Guide 25 -is there only one way?, in: Proceedings NCSL Workshop & Symposium, Boulder, CO, USA, 1998.
  47. ISO, ISO 5725-1. Accuracy (Trueness and Precision) of Measurement Methods and Results-General Principles and Definitions, International Organization for Standardization, Geneva, 1994.
  48. A. Rao, G.L. Lohse, Towards a texture naming system: identifying relevant dimensions of texture, Vision Research 36 (11) (1996) 1649-1669.
  49. J.R. Beveridge, K. She, B. Draper, G.H. Givens, Parametric and non-parametric methods for the statistical evaluation of humanid algorithms, in: Proceedings of the IEEE CVPR Workshop on Empirical Evaluation in Computer Vision, Kauai, HI, USA, 2001.
  50. B. Efron, R.J. Tibshirani, An Introduction to the Bootstrap, Chapman & Hall, London, 1993.
  51. A.P. Araujo, Analysis of variance of primary data on plant growth analysis, Pesquisa Agropecuária Brasileira 38 (2003) 1-10.
  52. E. Zappa, G. Busca, Comparison of eight unwrapping algorithms applied to fourier-transform profilometry, Optics and Lasers in Engineering 46 (2) (2008) 106-116.
  53. Z. Zhang, Determining the epipolar geometry and its uncertainty: a review, International Journal of Computer Vision 27 (2) (1998) 161-195.
  54. S. Zhang, Recent progresses on real-time 3d shape measurement using digital fringe projection techniques, Optics and Lasers in Engineering 48 (2) (2010) 149-158.
  55. J. Salvi, J. Pagès, J. Batlle, Pattern codification strategies in structured light systems, Pattern Recognition 37 (4) (2004) 827-849.
  56. C. Chen, A. Kak, Modeling and calibration of a structured light scanner for 3-d robot vision, in: Proceedings of the IEEE International Conference on Robotics and Automation, vol. 4, 1987, pp. 807-815. http://dx.doi.org/10.1109/ROBOT. 1987.1087958.
  57. G. Stein, Lens distortion calibration using point correspondences, in: Proceed- ings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1997, pp. 602-608. http://dx.doi.org/10.1109/CVPR.1997.609387.
  58. M. Trobina, Error Model of a Coded-Light Range Sensor, Technical Report BIWI-TR-164; ETH, Zürich, Switzerland, 1995.
  59. L.C. Ebert, W. Ptacek, S. Naether, M. Fürst, S. Ross, U. Buck, et al., Virtobot-a multi-functional robotic system for 3d surface scanning and automatic post mortem biopsy, International Journal of Medical Robotics and Computer Assisted Surgery 6 (1) (2010) 18-27.
  60. W. Schweitzer, M. Häusler, W. Bär, M. Schaepman, Evaluation of 3d surface scanners for skin documentation in forensic medicine: comparison of bench- mark surfaces, BMC Medical Imaging 7 (2007) 1.
  61. W. Schweitzer, M. Häusler, W. Bär, M. Schaepman, Retraction: evaluation of 3d surface scanners for skin documentation in forensic medicine: comparison of benchmark surfaces, BMC Medical Imaging 8 (2008) 15.