On the issue of variance depth weighted differences in the article CVPR2008

[360772864]

Thank you for your work. May I ask which software or website submitted the test results for the variance depth weighted difference graph in the article you cited (cvpr2008), as shown in the figure：

[pmoulon]

Can you remind us the article you refer too?

[360772864]

Thank you for your reply. It was in the code you provided that
：

SfM Camera trajectory quality evaluation

Those datasets can be used to evaluate the SfM camera trajectory quality found by
openMVG.

---

"dense multi-view stereo"

• http://cvlabwww.epfl.ch/data/multiview/
• C. Strecha, W. von Hansen, L. Van Gool, P. Fua, U. Thoennessen
  "On Benchmarking Camera Calibration and Multi-View Stereo for High Resolution Imagery"
  CVPR 2008.

Among them, this literature first evaluated the camera trajectory quality (i.e. the code you provided), and then calculated the variance depth weighted difference