Projects:2019s1-169 Forensic Applications of 3D Scanning

Project Team

Students

Jimmy Tang

Glenn Walsh

Supervisors

Dr Matthew Sorell

Richard Matthews

Introduction

With the advancement of technology, collecting and visualising physical evidence is inhibited by the limitations of traditional evidence collection, for example: photographs, casting shoe and footprints using plaster, and performing physical autopsies. Newer cutting edge technology allows for 3D models to be generated by capturing real world scenes can provide a more convenient method of collecting, and analysing 3D evidence. This can be done through 3D scanning shoe prints, crime scenes, and performing virtual autopsies, leading to more convenient and cost effective evidence collection. In this paper, we evaluate 3D imaging technology against traditional methodologies which are used in crime scene evidence collection.

We compare and contrast current 3D imaging techniques such as photogrammetry, laser scanning, and structured light systems. We expect, that although photogrammetry encounters limitations that 3D scanning doesn't, it still has its place as a lower cost method of visualising physical evidence. Due to the variety of evidence that can be collected, we focus our analysis on three main applications, that of foot prints, blood pattern collection and analysis, and digital autopsy.

In this paper we introduce existing consumer 3D scanning and photogrammetry tools and applications in order to investigate the 3D imaging of crime scenes. We evaluate the technologies effectiveness on three specific real world use cases. Specifically prints, blood patterns, and autopsy objects of interest. We will show whether 3D scanning and photogrammetry work in certain scenarios, are reliable, effective and how they compare to traditional methods such as physical measurements and high end laser scanning. We explore the limitations of the technologies and compare to the limitations of currently used forensic methodologies

Problem and Relevance

Going into the traditional methods of investigation, we focus our research on three specific real world applications where 3D scanning has the potential to introduce a step change. These areas are shoe or footprint impressions, blood pattern analysis, and autopsy. === Shoe and footprint impressions Shoe print impressions require a surface capable of holding impression; either sand, soil, or snow. Shoe impressions are easily obtained from these medium to to the ability of the material to keep their shape and form. Introducing a plaster cast ultimately introduces foreign material into the scene, thus has an effect on the investigation through evidence having been changed. Since the scene is contaminated through obtaining a cast, there is a risk that any cast made may contain subtle imperfections which may confuse an investigation. Consider also the fact that shoe prints may appear on any surface even non-impressionable surfaces such as shagged carpet of which are difficult to create create impressions of with traditional forensic methodologies. 3D scanning and imaging is a means to reconstruct the scene but without disturbing the evidence. 3D scanning also provides more versatility in the sense of being able to reconstruct non-impressionable surfaces as well. === Blood Pattern Analysis (BPA) Blood patterns give a useful indication where bleeding has occurred. Through trained analysis other useful information such as directionality and force can also be determined. Using traditional BPA, some useful forensic information can be inferred such as direction the victim was wounded, movements and positioning, causes of wounds (blunt trauma, gunshot, knife)[1]. The issue with traditional methods such as taking a photograph and performing blood pattern analysis is that there is room to misinterpret the information such the size of the patterns, lens distortion in photographs, and curved surfaces. 3D scanning and imaging allows for measurements, to be obtained more accurately by accounting for lens distortion, and because of the nature of 3D images, curved surfaces can be examined from any angle.

Autopsy

An Autopsy is a process of which to identify the cause of death and is also known as post-mortem examination. The value of autopsy is often seen negatively in terms of cost as the procedure is not reimbursable, fear of malpractice, along with legal and attitudinal issues often expressed but never proved. [2]. There are also religious implications to consider when doing a traditional autopsy. For example, some religions require the body to be buried within a certain time frame, and disallows traditional biopsy procedures. One such example of religious traditions is in Islam, Muslim traditions after death include burying the deceased as soon as possible, usually within 24 hours [3] This limitation gives reason to investigate 3D imaging techniques in autopsies to create a virtual autopsy environment. Technology behind 3D Scanning and Imaging

3D Scanning and Imaging

3D scanning and imaging are methods and techniques of extracting a 3D model out of physical objects such as equipment, walls, objects, foot prints, bodies, and textures. A traditional camera takes 2D images, and can be captured by many devices such as dedicated cameras, mobile phones, and surveillance systems. Bringing the technology to 3D images requires an additional depth parameter in order to describe where in 3D space this particular object exists in relation to its surroundings. Imaging surfaces deals with measurements of the (x, y, z) coordinates but more general 3D surface imaging systems associate each point on a non-planar surface a value such as surface reflectance [4]. This general 3D imaging system is referred to in the literature [4] as a point cloud and is made of 4 components P = (x, y, z ,f) where P represents the point cloud, x, y, z represent where a particular point is in 3D space, and f represents the value at this point; generally the translucency, or reflectivity.

Photogrammetry

Photogrammetry is a low cost, highly computationally expensive technique for creating a 3D model. This method of creating 3D models uses a series of photos to generate a point cloud of which can be texturised and then parsed into a 3D model using software. The model can be rendered on programs on 3D computer graphics software such as Blender.

Structured Light Systems

Structured Light Systems uses a projector and camera setup [8], where the projector can project a series of known light patterns over the scene, then use the properties of light reflecting and bending on a surface to reconstruct the scene as a 3D model. A limitation of this technique is that the rig must be able to project and view the pattern of light, this means that highly illuminated scenes will present challenges [8].

Preliminary Results

After investigating previous work, background knowledge, and current technology, we expect that photogrammetry will be inferior to 3D scanning in most aspects such as millimetre accuracy. However, these drawbacks are made up by being a low cost but still effective alternative to laser scanning. The overhead for current 3D scanning equipment is considered to be quite inhibiting, as such, photogrammetry offers an accessible way to obtain 3D data that overcomes the limitations of traditional policing methods. This isn’t to say that 3D scanning and photogrammetry don't have their own limitations, such as reflective or transparent surfaces. In these instances traditional methods will still prove to be a robust solution.

References

[1] N/A, “A simplied guide to bloodstain pattern analysis," National Forensic Science Technology Center R, Tech. Rep., N/A.

[2] The value of the autopsy in three medical eras," The New England journal of medicine., vol. 308, no. 17, 1983.

[3] E. C. Burton and K. A. Collins, “Religions and the autopsy," Medscape News and Perspective, 2012.

[4] J. Geng, “Structured-light 3d surface imaging: a tutorial," Adv. Opt. Photon., vol. 3, no. 2, pp. 128{160, Jun 2011. [Online]. Available: http://aop.osa.org/abstract.cfm?URI=aop-3-2-128

[5] C. Danakis, M. Afgani, G. Povey, I. Underwood, and H. Haas, “Using a cmos camera sensor for visible light communication," in 2012 IEEE Globecom Workshops, Dec 2012, pp. 1244{1248.

[6] S. Colwill, “Low-cost crime scene mapping: Reviewing emerging freeware, low-cost methods of 3d mapping reviewing emerging freeware, low-cost methods of 3d mapping evidence," 2016.

[7] H. H. Bartholomeus. (2019) Msc thesis subject: Importance of camera calibration for uav-based photogrammetry. [Online]. Avail- able: https://www.wur.nl/en/article/MSc-thesis-subject-Importance-of- camera-calibration-for-UAV-based-photogrammetry.htm

[8] D. Scharstein and R. Szeliski, \High-accuracy stereo depth maps using struc- tured light," in 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings., vol. 1. IEEE, 2003