3-D Scanning : Pre-Lecture and Workshop ResearchPosted: January 27, 2015
Tomorrow I will be having 3-D scanning session with Ingrid Murphy. This post is covering my background research before that session. I hope to utilize scanning in my work and so I am interested as to what kinds of scanning there are, what’s out there, how I could use it, what’s possible and what sorts of software I could learn. I’m particularly interested in open-source software and applications.
My first port of call was Wikipedia for a short definition : “A 3D scanner is a device that analyses a real-world object or environment to collect data on its shape and possibly its appearance (e.g. colour). The collected data can then be used to construct digital three-dimensional models.”
“common applications of this technology include industrial design, orthotics and prosthetics,reverse engineering and prototyping, quality control/inspection and documentation of cultural artifacts.”
The rest of the wiki is here, it makes for interesting reading. I read through this to gain an understanding of different types, applications and pros and cons of 3D scanning technology.
Seeing as Wikipedia is …Wikipedia I decided to read around to just make sure my overall knowledge of 3D scanning is correct.
This is some information I found in a paper called: The 3D Model Acquisition Pipeline
The most common range scanners are triangulation
systems. A lighting system projects a pattern of light onto the object to be scanned—
possibly a spot or line produced by a laser, or a detailed
pattern formed by an ordinary light source passing through
a mask or slide. A sensor, frequently a CCD camera, senses
the reflected light from the object. Software provided with
the scanner computes an array of depth values, which can
be converted to 3D point positions in the scanner coordinate
systems, using the calibrated position and orientation of the
light source and sensor. The depth calculation may be made
robust by the use of novel optics, such as the laser scanning
systems developed at the National Research Council of
Canada . Alternatively, calculations may be made robust
by using multiple sensors . A fundamental limitation of
what can be scanned with a triangulation system is having
an adequate clear view for both the source and sensor
to see the surface point currently being scanned. Surface
reflectance properties affect the quality of data that can be
obtained. Triangulation scanners may perform poorly on
materials that are shiny, have low surface albedo, or that
have significant subsurface scattering.
An alternative class of range scanners are time-of-flight
systems. These systems send out a short pulse of light, and
estimate distance by the time it takes the reflected light
to return. These systems have been developed with near
real time rates, and can be used over large (e.g. 100 m)
distances. Time-of-flight systems require high precision in
time measurements, and so errors in time measurement
fundamentally limit how accurately depths are measured.
Basic characteristics to know about a range scanner are
its scanning resolution, and its accuracy.
Accuracy is a statement of how close the measured value is to the true
value. The absolute accuracy of any given measurement is
unknown, but a precision that is a value for the standard
deviation that typifies the distribution of distances of
the measured point to true point can be provided by the
Resolution is the smallest distance between two points
that the instrument measures. The accuracy of measured 3D
points may be different than the resolution. For example, a
system that projects stripes on an object may be able to find
the depth at a particular point with submillimeter accuracy.
However, because the stripes have some width, the device may only be able to acquire data for points spaced millimetres apart on the surface. Resolution provides a fundamental bound on the dimensions of the reconstructed surface elements, and dictates the construction of intermediate data structures used in forming the integrated representation.
Note: For all but the simplest objects, multiple range scans must be acquired to cover the whole object’s surface. The individual range images must be aligned, or registered, into a common coordinate system so that they can be integrated into a single 3D model.
On this website I found a nice brief breakdown of the different types of scanners.
Note: Many systems rely on interactive alignment: a human operator is shown side-byside views of two overlapping scans, and must identify three or more matching feature points on the two images which are used to compute a rigid transformation that aligns the points.
This scanner is, as its name suggests, based on standard photography practices. The system takes multiple images of the object using reference points from each differing angle from which images are taken producing scan data. Please see our compressor cab casestudy (right).