Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Methodology for evaluation of precision and accuracy of different geometric 3D data acquisition methods

 
: Do Nascimento Melo, Ana Carolina
: Jahnke, Matthias; Knippers, Richard; Domajnko, Matevz

München, 2017, 84 S.
München, TU, Master Thesis, 2017
Englisch
Master Thesis
Fraunhofer IGD ()
Guiding Theme: Digitized Work; Research Area: Computer graphics (CG); Research Area: Computer vision (CV); 3D scanning; uncertainty visualization; guideline; standard; quality estimation

Abstract
3D optical scanning systems have been gaining considerable space in metrology, being largely applied in industry sectors and in the cultural heritage domain. The amount of available sensors on the market has grown considerably. Thereby, deciding for the right technique that fits-to-a-purpose or the most cost efficient technology, is a challenging task. When deciding in which technology to invest, the user often relies on the manufacturer’s instructions. However, manufacturers generally do not state under which conditions such values were acquired and thus, the system’s reproducibility is not assured. If measurements could be traced back to a common standard, this problem could be easily addressed. As such a solution is still not available, specialist often tend to solve this issue by associating terms like precision, accuracy and uncertainty to a measurement. Nowadays, the most applicable solution to define the accuracy of a system relies on the VDI/VDE 2634. This master thesis aims to develop a common solution to assess accuracy for different geometric 3D data acquisition models, considering the specifications of the VDI/VDE 2634 Part 3. The methodology proposed here encompasses the entire process from the acquisition to its processing stage. The study-case comprehend triangulated methods, as photogrammetry and laser line sensor. During the acquisition, a calibrated probing body and adapted test are proposed. The processing stage includes a best-fit algorithm and an evaluation of measurement uncertainty. The result comprehends the quality parameters together with the visualization of measurement uncertainty supporting the entire system. Therefore, providing to the end user enough information about the capability of the evaluated system.

: http://publica.fraunhofer.de/dokumente/N-509760.html