Statistical features of speckle-patterns for stereophotogrammetric 3D-reconstruction

Andreas Walter Stark, Eugene Wong, Holger Babovsky, Richard Kowarschik
  1. Institute of Applied Optics and Biophysics, Friedrich Schiller University Jena

andreas [dot] stark [at] uni-jena [dot] de

In an active stereophotogrammetric setup sinusoidal, bandlimited or speckle-patterns are projected sequentially onto an object to find a time based correlation. Out of these three only the speckle-patterns have a high depth of field and allow at the same time a high speed pattern generation. However their statistics limit their application compared to, for example, band limited patterns. The histogram of a speckle-distribution shows a clear concentration of intensities around an average value. In this work, we present a technique for speckle-based setups to decrease this disadvantageous effect by frequency filtering of the pattern and adapting the reconstruction algorithm to the statistical features of speckle-patterns. Possible advantages and disadvantages for the performance are investigated.

Manuscript not yet submitted. The submission phase is currently closed.
@inproceedings{stark2016b27,
  author     = {Stark, Andreas Walter and Wong, Eugene and Babovsky, Holger and Kowarschik, Richard},
  title      = {{Statistical features of speckle-patterns for stereophotogrammetric 3D-reconstruction}},
  booktitle  = {DGaO-Proceedings 2016},
  series     = {DGaO-Proceedings},
  eventtitle = {117. Jahrestagung der DGaO},
  eventdate  = {2016-05-18/2016-05-21},
  venue      = {Hannover},
  publisher  = {Deutsche Gesellschaft für angewandte Optik e.V.},
  year       = {2016},
  pages      = {B27},
  url        = {https://dgao-proceedings.de/paper/117-b27}
}
117. Annual Conference of the DGaO · Hannover · 2016