scan_matching
所属分类:matlab编程
开发工具:matlab
文件大小:2603KB
下载次数:241
上传日期:2011-08-14 10:58:38
上 传 者:
wolfeiwang
说明: 悉尼大学博士TimBaily编写的激光测距仪匹配的程序。其中包含C++的实现,内容包括:最小二乘优化,近邻匹配等基本算法的实现,是应用激光测距仪进行定位和地图构建不可多得的入门教程。
(scan-matching code written by matlab and c++, dealing with laser range data.)
文件列表:
scan_matching\correlate_variance (0, 2004-11-18)
scan_matching\correlate_variance\version1_gauss_svd (0, 2004-11-20)
scan_matching\correlate_variance\version1_gauss_svd\align_svd.m (1807, 2004-11-20)
scan_matching\correlate_variance\version1_gauss_svd\data.mat (6096, 2004-11-20)
scan_matching\correlate_variance\version1_gauss_svd\demo_gauss_svd.m (1130, 2004-11-20)
scan_matching\correlate_variance\version1_gauss_svd\gauss_variance_svd.m (857, 2004-09-09)
scan_matching\correlate_variance\version1_gauss_svd\make_gaussian_sum.m (730, 2004-09-07)
scan_matching\correlate_variance\version1_gauss_svd\sigma_ellipse.m (331, 2004-10-20)
scan_matching\correlate_variance\version1_gauss_svd\sog_correlate.dll (65536, 2004-11-19)
scan_matching\correlate_variance\version1_gauss_svd\sog_optimise.dll (73728, 2004-11-19)
scan_matching\correlate_variance\version1_gauss_svd\transform_to_global.m (479, 2004-10-21)
scan_matching\correlate_variance\version2_importance_sampling (0, 2004-11-20)
scan_matching\correlate_variance\version2_importance_sampling\data.mat (6096, 2004-11-20)
scan_matching\correlate_variance\version2_importance_sampling\demo_importance_gauss.m (1172, 2004-11-20)
scan_matching\correlate_variance\version2_importance_sampling\importance_gaussian.m (1097, 2004-11-20)
scan_matching\correlate_variance\version2_importance_sampling\make_gaussian_sum.m (730, 2004-09-07)
scan_matching\correlate_variance\version2_importance_sampling\multivariate_gauss.m (388, 2004-10-22)
scan_matching\correlate_variance\version2_importance_sampling\sample_mean.m (842, 2002-02-11)
scan_matching\correlate_variance\version2_importance_sampling\sigma_ellipse.m (331, 2004-10-20)
scan_matching\correlate_variance\version2_importance_sampling\sog_correlate.dll (65536, 2004-11-19)
scan_matching\correlate_variance\version2_importance_sampling\stratified_random.m (337, 2003-12-05)
scan_matching\correlate_variance\version2_importance_sampling\stratified_resample.m (790, 2004-10-20)
scan_matching\correlate_variance\version2_importance_sampling\transform_to_global.m (479, 2004-10-21)
scan_matching\least_squares_align (0, 2004-11-17)
scan_matching\least_squares_align\align_least_squares.m (2457, 2004-11-17)
scan_matching\least_squares_align\data.mat (1319672, 2004-11-16)
scan_matching\least_squares_align\demo_least_squares.m (2234, 2004-11-19)
scan_matching\least_squares_align\pi_to_pi.m (569, 2004-07-19)
scan_matching\least_squares_align\transform_to_global.m (479, 2004-10-21)
scan_matching\mexdemos (0, 2004-11-19)
scan_matching\mexdemos\demo_icp.m (1228, 2004-11-19)
scan_matching\mexdemos\demo_icp_2.m (1654, 2004-11-19)
scan_matching\mexdemos\demo_icp_3.m (1285, 2004-11-19)
scan_matching\mexdemos\demo_k_neighbours.m (805, 2004-11-18)
scan_matching\mexdemos\demo_nearest_neighbour.m (682, 2004-11-18)
scan_matching\mexdemos\demo_sog_correlate.m (4641, 2004-11-19)
scan_matching\mexdemos\demo_sog_optimise.m (1118, 2004-11-19)
scan_matching\mexdemos\icp.dll (65536, 2004-11-19)
scan_matching\mexdemos\k_neighbours.dll (61440, 2004-11-18)
scan_matching\mexdemos\laserdata.mat (7558752, 2002-01-21)
... ...
Scan Matching, Tim Bailey 2004.
-------------
This collection of MatLab scripts, mex-functions and C++ code is a set
of code for aligning scans of 2D point data. In particular, it is designed
to facilitate matching of laser-scans.
The directories and their contents are listed below. The first 3 are related
and contain code and demos of scan-matching mex-files. The last 2 contain
MatLab scripts: the former has m-files for computing scan-match covariances,
and the latter has m-files for batch least-squares alignment of a set of
scan-match estimates.
mexdemos
--------
This directory has a number of files prefixed "demo" which demonstrate the
usage of the scan-matching mex-files. These demos are reasonably well
documented and self-explanatory.
They show how to perform iterated closest point (ICP) scan alignment and
Sum of Gaussian (SoG) correlation. They also show some 2D nearest-neighbour
routines.
mexfiles
--------
This directory contains the source code for all the mex-files. There is
also a MS Visual C++ 7 project for compiling them to mex-dlls.
source
------
This directory contains the set of C++ source modules that provide all
the functionality for the mex-code in the "mexfiles" directory.
correlate_variance
------------------
This directory has two sets of m-files for computing the covariance between
two aligned scans. The basic idea is, given an estimate of the relative pose between
two scans, we would like to compute a Gaussian covariance of this pose estimate.
The first algorithm is deterministic, and tries to fit a Gaussian function to
samples from the correlation function. It uses SVD to cope with an overconstrained
system of equations.
The second algorithm is Monte Carlo and estimates the variance of the correlation
function using importance sampling.
least-squares align
-------------------
This directory has a set of m-files to perform least-squares alignment of
relative pose estimates (as obtained from scan-matching). It implements a
simple iterated non-linear least-squares algorithm. The demo program
shows how this code works.
近期下载者:
相关文件:
收藏者: