We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Comparison of Depth Estimation Setups from Stereo Endoscopy and Optical Tracking for Point Measurements

Abstract: To support minimally-invasive intraoperative mitral valve repair, quantitative measurements from the valve can be obtained using an infra-red tracked stylus. It is desirable to view such manually measured points together with the endoscopic image for further assistance. Therefore, hand-eye calibration is required that links both coordinate systems and is a prerequisite to project the points onto the image plane. A complementary approach to this is to use a vision-based endoscopic stereo-setup to detect and triangulate points of interest, to obtain the 3D coordinates. In this paper, we aim to compare both approaches on a rigid phantom and two patient-individual silicone replica which resemble the intraoperative scenario. The preliminary results indicate that 3D landmark estimation, either labeled manually or through partly automated detection with a deep learning approach, provides more accurate triangulated depth measurements when performed with a tailored image-based method than with stylus measurements.
Comments: Accepted at Bildverarbeitung fuer die Medizin (BVM), Informatik aktuell. Springer Vieweg, Wiesbaden 2022
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2201.10848 [cs.CV]
  (or arXiv:2201.10848v1 [cs.CV] for this version)

Submission history

From: Lalith Sharan [view email]
[v1] Wed, 26 Jan 2022 10:15:46 GMT (3157kb,D)

Link back to: arXiv, form interface, contact.