rob_hess_pf
所属分类:图形图像处理
开发工具:Visual C++
文件大小:19989KB
下载次数:600
上传日期:2009-12-18 10:51:17
上 传 者:
zhengyufeng2009
说明: 粒子滤波现在已经成为目标跟踪领域的主流算法,它的应用范围广泛,在非线性、非高斯噪声下依然表现良好。该代码是Rob Hess 编写的。他的个人主页是:http://web.engr.oregonstate.edu/~hess/
(Now,Partical filter has become the main algorithm in moving target tracking region.It still perform very well in nonlinear non-gaussian noise conditions. So its application is wide.
The code is written by Rob Hess.His homeage is:http://web.engr.oregonstate.edu/~hess/)
文件列表:
rob hess PF\ROB HESS PF track ppt.pdf (11856023, 2009-12-03)
rob hess PF\track\bin\Makefile (797, 2006-03-19)
rob hess PF\track\Doxyfile (45621, 2006-03-23)
rob hess PF\track\hist.plt (204, 2006-03-20)
rob hess PF\track\include\defs.h (931, 2006-03-11)
rob hess PF\track\include\observation.h (4618, 2006-03-20)
rob hess PF\track\include\particles.h (3870, 2006-03-19)
rob hess PF\track\include\utils.h (3886, 2006-03-08)
rob hess PF\track\mainpage (132, 2006-03-23)
rob hess PF\track\Makefile (384, 2006-03-23)
rob hess PF\track\src\Makefile (477, 2006-03-14)
rob hess PF\track\src\observation.c (6971, 2006-03-20)
rob hess PF\track\src\observe.c (4133, 2006-03-23)
rob hess PF\track\src\particles.c (5480, 2006-03-20)
rob hess PF\track\src\track1.c (12310, 2006-03-23)
rob hess PF\track\src\utils.c (5698, 2006-03-14)
rob hess PF\track\video\hockey.avi (5550554, 2006-03-19)
rob hess PF\track\video\soccer.avi (4353388, 2006-03-19)
rob hess PF\track\bin (0, 2009-12-03)
rob hess PF\track\docs (0, 2009-12-03)
rob hess PF\track\frames (0, 2009-12-03)
rob hess PF\track\include (0, 2009-12-03)
rob hess PF\track\src (0, 2009-12-03)
rob hess PF\track\video (0, 2009-12-03)
rob hess PF\track (0, 2009-12-16)
rob hess PF (0, 2009-12-17)
Contents:
1. Intro
2. Compiling
3. Running
a. track1
b. observe
4. Histograms
/***********************************************************************/
1. Intro
This distribution contains source code for a simple object tracking
program using particle filtering. You need to have the OpenCV Library
and the GNU Scientific Library (GSL) installed to compile and use the
programs. See the below two links:
http://sourceforge.net/projects/opencvlibrary/
http://www.gnu.org/software/gsl/
/***********************************************************************/
2. Compiling
Some of the source files can be documented with Doxygen:
> make docs
To comile the executable programs:
> make bin
To compile everything (executables, documentation, etc.):
> make
To clean up:
> make clean
To get rid of executables, too:
> make distclean
/***********************************************************************/
3. Running
3.a. track1: Tracks a single object in a sequence of video.
Supply track1 with a video file on the command line:
> ./bin/track1 video.mpg
You will be presented with the first frame of the video, in which you are
required to select an object to track. Do this by defining a rectangular
region with the mouse by clicking and dragging. After you select the
object to be tracked, you will see the results of tracking. This usually
runs in about real-time, unless the region you've selected is too big.
You can get additional help for command line options and arguments by
passing the '-h' option:
> ./bin/track1 -h
3.b. observe: Displays the likelihood of an object at each pixel location
in an image.
Supply observe with an image in which to observe and a set of images from
which to compute a reference histogram upon which observation is based:
> ./bin/observe [ ...]
You will be shown an image representing the likelihood of the object
defined by the reference images at every pixel location in the image being
observed. You can get additional help for command line options and
arguments by passing the '-h' option:
> ./bin/observe -h
/***********************************************************************/
4. Histograms
You can choose to export the frames of video tracked by track1 using the
'-o' flag. If you choose to do so, the values of the reference histogram
are also output to the file hist_00.dat. The included file hist.plt is a
gnuplot script which can be used to visualize the reference histogram.
Use it as follows:
> gnuplot hist.plt
It will create an encapsulated postscript histogram image in hist.eps.
近期下载者:
相关文件:
收藏者: