Evaluating local features in high-resolution 3D-PLI data

Sharib Ali, Markus Axer, Katrin Amunts, Roland Eils, Karl RohrISBI 2018 · OpenReview

Abstract

Even though imaging techniques enhance some local features in histological brain sections, it is often difficult to detect these features. In this paper, we introduce a robust feature detection method for accurate registration of multi-resolution 3D polarized light imaging data of brain sections. Our approach is based on the structure tensor and a scale-space scheme that reduces ambiguity and preserves scale-invariant features. We applied our approach to data from tissue sections of human and rat brain. Quantitative evaluations are performed and compared with state-of-the-art methods.
BibTeX
@inproceedings{ali2018evaluating,
  author = {Sharib Ali and Markus Axer and Katrin Amunts and Roland Eils and Karl Rohr},
  title = {{Evaluating local features in high-resolution 3D-PLI data}},
  booktitle = {ISBI},
  year = {2018},
  url = {https://openreview.net/forum?id=wC2cu88-GYf},
}

Loading related content...

Figure