siftpp-0.7.3
所属分类:GDI/图象编程
开发工具:Visual C++
文件大小:352KB
下载次数:29
上传日期:2009-01-30 14:14:19
上 传 者:
幽灵王子
说明: 关于sift的编程,基于vc++,快速的实现特征匹配
(With regard to sift programming, based on vc++, the rapid realization of feature matching)
文件列表:
siftpp-0.7.3\benchmark.cpp (4369, 2007-03-23)
siftpp-0.7.3\data\img3.pgm (512015, 2006-08-22)
siftpp-0.7.3\doxygen.conf (46852, 2006-09-02)
siftpp-0.7.3\Makefile (2790, 2007-11-19)
siftpp-0.7.3\man.html (10905, 2006-10-28)
siftpp-0.7.3\notice.txt (1920, 2006-08-22)
siftpp-0.7.3\sift-conv.tpp (6293, 2006-11-04)
siftpp-0.7.3\sift-driver.cpp (24491, 2007-11-19)
siftpp-0.7.3\sift.1 (7639, 2007-03-29)
siftpp-0.7.3\sift.cpp (42805, 2007-11-14)
siftpp-0.7.3\sift.hpp (8790, 2007-04-04)
siftpp-0.7.3\sift.html (10903, 2006-10-28)
siftpp-0.7.3\sift.ipp (9866, 2007-04-04)
siftpp-0.7.3\siftpp.dsp (4524, 2008-03-25)
siftpp-0.7.3\siftpp.dsp.bak (4323, 2008-03-25)
siftpp-0.7.3\siftpp.dsw (536, 2008-03-25)
siftpp-0.7.3\siftpp.ncb (50176, 2008-03-25)
siftpp-0.7.3\siftpp.opt (125952, 2008-03-25)
siftpp-0.7.3\siftpp.plg (7697, 2008-03-25)
siftpp-0.7.3\test_atan2.m (601, 2006-08-22)
siftpp-0.7.3\TIMESTAMP (57, 2007-11-19)
siftpp-0.7.3\data (0, 2008-04-02)
siftpp-0.7.3\Debug (0, 2008-04-02)
siftpp-0.7.3 (0, 2008-04-02)
A LIGHTWEIGHT C++ IMPLEMENTATION OF
DAVID LOWE'S SCALE INVARIANT FEATURE TRANSFORMS
Andrea Vedaldi
http://www.cs.ucla.edu/~vedaldi
COPYRIGHT WARNING
This code implements an algorithm (SIFT) patended by David Lowe and
cannot be used for commercial purposes without his permission. See
the included license notice for more information.
CHANGES
0.7.3 Another bug fix introduced in 0.7.1.
0.7.2 Fixed a bug introduced with 0.7.1.
0.7.1 Tuned a few parameters to better match D. Lowe.
0.7.0 --unnormalized, --floating-point, --magnif
Bug fixes (thanks to Brian).
0.6.7 Bug fixes (thanks to Loic).
0.6.6 Bug fixes (thanks to Giuseppe).
0.6.5 Bug fixes (thanks to Liu).
0.*** Added -I. to Makefile.
0.6.3 Minor bug fixes (thanks to Erwin).
0.6.2 Minor bug fixes. Minor corrections to the man page.
0.6.1 Removed a bug affecting the downsampling of oddly sized images.
0.6.0 Introduced --stable-order option.
0.5.6 Fixed a minor bug in the convolution code. Increased the default
number of octaves that get computed.
0.5.5 More bug fixes (PGM parsing error)
0.5.4 New switch VL_LOWE_STRICT. This enables features that simulate
even more closely Lowe's implementation (it does even
simulate a small 'bug').
0.5.3 Bug fixes (thanks to P. Ivanov).
0.5.2 Man page.
0.5.1 Changed the way image boundaries are handled in the computation
of the scalespace (thanks to Q.S.Luo for indicating the
problem). This makes the algorithm even closer to
Lowe's. Inital version of the manpage.
0.5 Optionally reads keypoint frames from file. Improved interface.
0.4 Faster (and approximated) computations. Binary descriptor output.
0.3 Added `--threshold' and `--edge-threshold' opitons to the
command line driver.
0.2.2 Corrected a bug in the resolution of the 3x3 linear system.
(thanks to Fan Gu for finding this out)
0.2.1 Corrected a bug in the command line driver.
0.2 Made LAPACK independent.
0.1 Initial version.
近期下载者:
相关文件:
收藏者: