this is for holding javascript data
Volker Strobel edited section_Pillar_I_Training_Set__.tex
almost 8 years ago
Commit id: 8d25fe91c933389ec7d19f994fd6b1e27ae8b798
deletions | additions
diff --git a/section_Pillar_I_Training_Set__.tex b/section_Pillar_I_Training_Set__.tex
index 46171d5..8cefad0 100644
--- a/section_Pillar_I_Training_Set__.tex
+++ b/section_Pillar_I_Training_Set__.tex
...
move to different environments.
As an alternative, we sought a low-budget and more flexible
solution. While the homography-based approach
(Section xxx) (Section~\ref{sec:keypointmatching}) shows high matching quality
for non-blurry images, it suffers from lacking robustness to noise
and the high processing time requirements. To exploit the advantages,
and straighten out the disadvantages, we used the approach in a preprocessing