HOG-descripoter
所属分类:其他书籍
开发工具:C/C++
文件大小:1146KB
下载次数:16
上传日期:2012-12-04 14:58:05
上 传 者:
hengwings
说明: HOG论文提供的源代码,非常不错 值得借鉴
(HOG papers provide the source code, very good worth learning)
文件列表:
HOG descripoter\HOG\aclocal.m4 (262947, 2008-10-16)
HOG descripoter\HOG\app\classify_rhog.cpp (2880, 2009-03-07)
HOG descripoter\HOG\app\dump4svmlearn.cpp (6934, 2009-03-07)
HOG descripoter\HOG\app\dumpsegd.cpp (9728, 2008-10-16)
HOG descripoter\HOG\app\dump_rhog.cpp (2757, 2009-03-07)
HOG descripoter\HOG\app\formatstream.h (11070, 2007-06-12)
HOG descripoter\HOG\app\Makefile.am (1030, 2008-10-16)
HOG descripoter\HOG\app\Makefile.in (24001, 2008-10-16)
HOG descripoter\HOG\app\rawdescio.cpp (4601, 2009-03-07)
HOG descripoter\HOG\app\rawdescio.h (5703, 2009-03-07)
HOG descripoter\HOG\app\segobj.h (13971, 2006-05-22)
HOG descripoter\HOG\app\test_library.cpp (2350, 2009-03-07)
HOG descripoter\HOG\app\windetectmain.cpp (14230, 2009-03-07)
HOG descripoter\HOG\app\windetectmain.h (5269, 2009-03-07)
HOG descripoter\HOG\AUTHORS (1, 2007-06-12)
HOG descripoter\HOG\autom4te.cache\output.0 (819944, 2008-09-05)
HOG descripoter\HOG\autom4te.cache\output.1 (819374, 2008-09-05)
HOG descripoter\HOG\autom4te.cache\output.2 (802131, 2008-10-16)
HOG descripoter\HOG\autom4te.cache\output.3 (801561, 2008-10-16)
HOG descripoter\HOG\autom4te.cache\requests (24813, 2008-10-16)
HOG descripoter\HOG\autom4te.cache\traces.0 (482678, 2008-09-05)
HOG descripoter\HOG\autom4te.cache\traces.1 (57285, 2008-09-05)
HOG descripoter\HOG\autom4te.cache\traces.2 (471732, 2008-10-16)
HOG descripoter\HOG\autom4te.cache\traces.3 (58304, 2008-10-16)
HOG descripoter\HOG\bootstrap (910, 2008-09-05)
HOG descripoter\HOG\buildtest.sh (976, 2007-06-14)
HOG descripoter\HOG\ChangeLog (1, 2007-06-12)
HOG descripoter\HOG\cleanall (249, 2007-06-12)
HOG descripoter\HOG\config.guess (44878, 2008-09-05)
HOG descripoter\HOG\config.h.in (1855, 2008-09-05)
HOG descripoter\HOG\config.sub (33387, 2008-09-05)
HOG descripoter\HOG\configure (800445, 2008-10-16)
HOG descripoter\HOG\configure.in (5674, 2008-09-05)
HOG descripoter\HOG\COPYING (17993, 2007-06-12)
HOG descripoter\HOG\depcomp (17867, 2008-09-05)
HOG descripoter\HOG\Doxyfile (10185, 2007-06-12)
HOG descripoter\HOG\INSTALL (9127, 2007-06-12)
HOG descripoter\HOG\install-sh (13620, 2008-09-05)
HOG descripoter\HOG\lear\blitz\blitzio.h (7947, 2007-06-12)
HOG descripoter\HOG\lear\blitz\blitzutil.h (5350, 2007-06-12)
... ...
Author: Navneet Dalal (navneet.dalal@inrialpes.fr)
=== IMPORTANT NOTE ============================================================
I have limited capabilites for supporting the code. Hence the code below is
provided as it is without any support.
If you find some bugs, you may report to me and I will fix them. Even better,
send the files you have fixed and I will upload them directly.
=== WINDOWS SUPPORT ==========================================================
I have received lots of emails regarding providing the binaries/source for
windows. Due to time-constraints, I will not be able to provide windows
source/binaries. However if are really interested, it is not hard to compile
it for windows.
All you need to do is
1. Change image IO part in lear/image/imageio.h such that instead of using
imlib2, it uses your image io libraries.
2. Fix painting classes lear/image/painter.h and lear/image/pimage.h such
that it uses your painting routines.
If you have created MS Visual C++ workspace, please send those files to
me so that I can include them in the package.
=== SVM LEARN FILES ==========================================================
This code does not build svm_learn binary. For that download
http://pascal.inrialpes.fr/soft/olt/svmdense.tar.gz
and build a svm_learn binary. A pre-built binary is already provided in
the binary package.
=== GENERAL INFORMATION =======================================================
This code implements a multiscale detector for visual object
recognition and notably for human detection. It requires three external
libraries:
-- Imlib2 for reading/writing images
-- Blitz for efficient array processing
-- Boost for many other general c++ utilities not present in ISO C++
Some portions of the code (notably the color conversion routines)
contain sources adapted from the Vigra Image Processing library
(http://kogs-www.informatik.uni-hamburg.de/~koethe/vigra/). Vigra is
templated image processing library distributed under the MIT X11
license. Where ever Vigra code has been used, this is mentioned in the
corresponding header file. A license file is included in the
distribution.
=== INSTALLATION ==============================================================
First install the development versions (*-devel RPM's or similar) of
the three external libraries as follows, making sure that the headers
and libraries are in your standard search paths :
Imlib2 - version 1.2.2 tested
Boost - version 1.33.1 or later
Blitz - version 0.9 tested (installed from source as no RPM's are
currently available)
Then unpack the learcode.tar.gz file, change to it top level directory
and run ./configure. If the libraries were installed in nonstandard
places you can tell configure where to find them using, e.g.
./configure --with-blitz=/usr/local --with-boost=/usr/local ...
=== LICENSE ===================================================================
This software is distributed under the "CeCILL-B FREE SOFTWARE LICENSE
AGREEMENT" detailed in the LICENSE file, and subject to the additional
requirement (Article 14) regarding acknowledgements. The following
scientific publications contain the original descriptions of the
method implemented in this Software:
[1] Histograms of Oriented Gradients for Human Detection. Navneet
Dalal and Bill Triggs. IEEE Conference on Computer Vision & Pattern
Recognition (CVPR), vol. II, pages 886-893, June 2005.
[3] Finding people in images and videos. Navneet Dalal. PhD thesis,
Institut National Polytechnique de Grenoble, July 2006.
近期下载者:
相关文件:
收藏者: