3D scanning and visualization for crime scenes
3D scanning and visualization for crime scenes
By: Tom Rataj
Back in the “old” days, serious crime scenes were documented with a few photographs taken with a medium or large format camera, using black and white print film and a single-use flash-bulb, when extra light was required. This was eventually replaced with 35mm cameras using colour print film and multi-use battery-powered electronic flashes.
In addition to photographs, more serious crime-scenes were also methodically measured and dimensioned and a skilled forensic officer drew blueprint-style plans, so the scene could later be studied, analysed and presented in court. All of this was of course extremely labour and time intensive and produced only 2-dimensional (2D) images.
Around 1990 still-photography was sometimes augmented by video walk-throughs of scenes, using analogue video-tapes.
In the early 2000s, much of the photography transitioned from colour-print film to digital, greatly reducing the cost of photography and improving the ability to quickly share imagery. Digital video cameras using solid-state memory cards also later replaced tape based video cameras, improving quality and the ability to quickly share video.
Manually drawn crime-scene plans also went digital with the advent of electronic drafting programs such as AutoCAD and others. Once the electronic plans were completed, large wide-format printers could quickly and accurately produce blue-print sized drawings within a few minutes.
While all these advances certainly revolutionised crime-scene imaging and mapping, and saved substantially amounts of time and labour, they still all relied on crime-scenes being measured with measuring tapes and other devices. Obviously, this presented numerous problems with speed and accuracy and introduced the potential for transpositional errors, all of which could undermine investigations and court proceedings.
Fortunately, there is another technology available that substantially advances crime-scene imagery. Laser-based scanning equipment from market leader Leica GeoSystems, which has been in use by law enforcement forensics and collision reconstruction investigations for several years, has now advanced to the point where it can rapidly and simultaneously capture 3D and high dynamic range (HDR) photos of any scene.
Their latest generation Leica ScanStation P40 can scan a scene at a scan rate of 1 million points per second at a range of up to 270m. It has high angular accuracy, low range noise and survey-grade dual-axis compensation so it is capable of creating highly accurate 3D colour point images with realistic clarity. The integrated digital camera can simultaneously capture both video and still images, offering additional versatility.
The P40 is also highly temperature tolerant (–20°C to +50°C / –4°F to 122°F) so it can function in even the harshest conditions.
How it works
The ScanStation uses a precisely focused laser which is aimed at the surface of objects within its range. The laser emits a pulse of light towards an object and then measures the amount of time it takes for the reflection of that pulse of light to return to the system. Since the speed of light is known, the time of the light-pulse’s round-trip produces a precise measurement of the distance between the ScanStation and the object(s) that laser pulse reaches.
To create a virtual image of an object or a crime scene, the 3D laser uses a system of rotating mirrors to aim the laser light in very fine incremental steps within the scanning area. Each neighbouring point of objects in the crime scene is measured and recorded by the scanner, and once all the measured points are assembled or reconstructed, the shapes of the scanned objects are displayed in what is called a point-cloud.
The visualization or picture produced by the 3D scanner also provides precise distance measurements between points on the surface of the object so users can determine the length, width and thickness of an object as well as the distance between any points in the scan. Reconstructing all the point-cloud information is done through a reference system in a process known as alignment or registration. To maintain a high level of accuracy the ScanStation is typically mounted on a commercial grade tripod which provides a stable, fixed position.
To get a complete 3D image of a room or other crime scene, the 3D scanner typically needs to be repositioned to different locations within the crime scene so that all objects are scanned from multiple sides. In the reference system, the scan information from each location is electronically stitched together to create a virtual 3D model of the scanned scene. The virtual 3D model is displayed on a monitor where it can be moved and otherwise manipulated to better study a scene.
Because the virtual reconstruction is so accurate, the size of objects and the distance between objects can be very accurately measured. In a crime scene for example, a precise measurement of the distance and orientation of an item of evidence can easily be determined relative to other objects in the scene such as furniture or walls.
The 3D scanner can also be used to accurately determine the field of view a person had from a particular location (such as a window or door), which would allow investigators to verify witness testimony or other information.
Further information is available at: http://www.leica-geosystems.us/en/index.htm
Tom Rataj is Blue Line’s editor and technology columnist. He retired from the Toronto Police Service in 2015 after 35 years of service. He can be contacted at email@example.com .