20 lines
849 B
Text
20 lines
849 B
Text
|
- 2010.04.08, Pierre Moulon
|
||
|
|
Use Eigen matrix library in order to compute the SVD in ORSA.
|
||
|
|
Add a Cmake based build, for easier compilation on Windows.
|
||
|
|
|
||
|
|
- 2010.11.08, Guoshen Yu
|
||
|
|
Remove a few minor bugs in orsa.cpp.
|
||
|
|
Code compiled with Intel C++ and Visual C++ compilers under Windows (32 bits). Both OpenMP and vectorization work with Intel C++. OpenMP but vectorization withs with Visual C++.
|
||
|
|
|
||
|
|
- 2010.08.17, Guoshen Yu
|
||
|
|
In compute_asift_matches.c, change the nfa_max value from 0 to -2 to reduce false matching.
|
||
|
|
|
||
|
|
- 2010.08.28, Guoshen Yu
|
||
|
|
In demo_lib_sift.cpp, change par.MatchRatio from 0.75 to 0.73 to reduce et number of false matches.
|
||
|
|
In demo_ASIFT.cpp, change the image resize target dimension from 640x480 to 800x600.
|
||
|
|
In demo_ASIFT.cpp, add a white band between the images in the output image for better visibility.
|
||
|
|
|
||
|
|
|
||
|
|
|
||
|
|
|