1）SLAM——using Kinect 2） SFM--CMVS PMVS Bundler 3）VXL

http://blog.exbot.net/archives/296

（2）http://www.di.ens.fr/cmvs/        http://download.csdn.net/detail/chlele0105/5607187#comment       http://www.cs.cornell.edu/~snavely/bundler/       http://blog.csdn.net/gisstar/article/details/6776041     http://blog.csdn.net/zzzblog/article/details/17166869

（3）http://vxl.sourceforge.net/

VXL (the Vision-something-Libraries)是计算机视觉研究和实现库集。它从TargetJr和IUE演变而来，目的是成为一个轻量级、速度快和持久的系统。它可移植到很多平台。 http://blog.csdn.net/houston11235/article/details/8146344    http://blog.csdn.net/stereohomology/article/details/27207505

包含的库
◆ 数字化容器和法则：vnl
◆ 图像管理：vil
◆ 几何图形：vgl
◆ I/O控制：vsl
◆ 基本模板：vbl
◆ 功能库：vul

项目主页：http://www.open-open.com/lib/view/home/1328931840421

最近打算使用Kinect实现机器人的室内导航，收集了近年来的一些比较好的文章。《基于Kinect系统的场景建模与机器人自主导航》、《Mobile Robots Navigation in Indoor Environments Using Kinect》、《Using a Depth Camera for Indoor Robot Localization and Navigation》、《Depth Camera Based Indoor Mobile Robot Localization and Navigation》、《Using a Depth Camera for Indoor Robot Localization and Navigation》、《Using the Kinect as a Navigation Sensor for Mobile Robotics》。

by Top Liu
最近打算使用Kinect实现机器人的室内导航，收集了近年来的一些比较好的文章。

基于Kinect系统的场景建模与机器人自主导航

《机器人》 2012年05

杨东方  王仕成  刘华平  刘志国  孙富春  

【摘要】：本文分别基于微软Kinect系统的单目RGB摄像机以及深度距离受限的RGB-D像机,研究解决室内机器人的6自由度定位问题.首先,在传统不完全自由度估计的基础上,提出了特征点参数的增量式模型以解决运动尺度不确定性问题.该模型和以往的欧几里得、逆深度参数化模型相比,不仅能够显著降低系统状态维数,而且能够保证系统状态的一致可观测性;此外,基于增量式模型,根据Kinect系统中采集的RGB图像和红外图像,实现了对机器人6自由度的运动估计.最后,将Kinect系统采集得到的RGB图像和深度图像序列用于欧几里得参数化模型和增量式参数化模型,对应的实验结果证明了本文所提的自主导航方法的有效性.

下载：http://robot.sia.cn/CN/article/downloadArticleFile.do?attachType=PDF&id=15382

 

国外文献：

1.Mobile Robots Navigation in Indoor Environments Using Kinect

This paper appears in:
Critical Embedded Systems (CBSEC), 2012 Second Brazilian 

没了方便没有IEEE账户的朋友，我已上传到百度文库。下面的文章可直接点击下载

2.Using a Depth Camera for Indoor Robot Localization and Navigation

 Abstract—Depth cameras are a rich source of information for
robot indoor localization and safe navigation. The recent availability
of the low-cost Kinect sensor provides a valid alternative
to other available sensors, namely laser-range finders. This
paper presents the first results of the application of a Kinect
sensor on a wheeled indoor service robot for elderly assistance.
The robot makes use of a metric map of the environment’s
walls and uses the depth information of the Kinect camera to
detect the walls and localize itself in the environment. In our
approach an error minimization method is used providing realtime
efficient robot pose estimation. Furthermore, the depth
camera provides information about the obstacles surrounding
the robot, allowing the application of path-finding algorithms
such as D* Lite achieving safe and robust navigation. Using
the proposed solution, we were able to adapt a robotic soccer
robot developed at the University of Aveiro to successfully
navigate in a domestic environment, across different rooms
without colliding with obstacles in the environment.

3.Depth Camera Based Indoor Mobile Robot Localization and Navigation

Abstract—The sheer volume of data generated by depth
cameras provides a challenge to process in real time, in
particular when used for indoor mobile robot localization and
navigation. We introduce the Fast Sampling Plane Filtering
(FSPF) algorithm to reduce the volume of the 3D point cloud
by sampling points from the depth image, and classifying local
grouped sets of points as belonging to planes in 3D (the “plane
filtered” points) or points that do not correspond to planes
within a specified error margin (the “outlier” points). We then
introduce a localization algorithm based on an observation
model that down-projects the plane filtered points on to 2D, and
assigns correspondences for each point to lines in the 2D map.
The full sampled point cloud (consisting of both plane filtered
as well as outlier points) is processed for obstacle avoidance
for autonomous navigation. All our algorithms process only
the depth information, and do not require additional RGB
data. The FSPF, localization and obstacle avoidance algorithms
run in real time at full camera frame rates (30Hz) with low
CPU requirements (16%). We provide experimental results
demonstrating the effectiveness of our approach for indoor
mobile robot localization and navigation. We further compare
the accuracy and robustness in localization using depth cameras
with FSPF vs. alternative approaches that simulate laser
rangefinder scans from the 3D data.

4.Using a Depth Camera for Indoor Robot Localization and Navigation

Abstract—Depth cameras are a rich source of information for
robot indoor localization and safe navigation. The recent availability
of the low-cost Kinect sensor provides a valid alternative
to other available sensors, namely laser-range finders. This
paper presents the first results of the application of a Kinect
sensor on a wheeled indoor service robot for elderly assistance.
The robot makes use of a metric map of the environment’s
walls and uses the depth information of the Kinect camera to
detect the walls and localize itself in the environment. In our
approach an error minimization method is used providing realtime
efficient robot pose estimation. Furthermore, the depth
camera provides information about the obstacles surrounding
the robot, allowing the application of path-finding algorithms
such as D* Lite achieving safe and robust navigation. Using
the proposed solution, we were able to adapt a robotic soccer
robot developed at the University of Aveiro to successfully
navigate in a domestic environment, across different rooms
without colliding with obstacles in the environment.

5.Using the Kinect as a Navigation Sensor for Mobile Robotics

ABSTRACT
Localisation and mapping are the key requirements in mobile
robotics to accomplish navigation. Frequently laser scanners
are used, but they are expensive and only provide 2D mapping
capabilities. In this paper we investigate the suitability
of the Xbox Kinect optical sensor for navigation and simultaneous
localisation and mapping. We present a prototype
which uses the Kinect to capture 3D point cloud data of the
external environment. The data is used in a 3D SLAM to
create 3D models of the environment and localise the robot
in the environment. By projecting the 3D point cloud into
a 2D plane, we then use the Kinect sensor data for a 2D
SLAM algorithm. We compare the performance of Kinectbased
2D and 3D SLAM algorithm with traditional solutions
and show that the use of the Kinect sensor is viable. However,
its smaller field of view and depth range and the higher
processing requirements for the resulting sensor data limit its
range of applications in practice.

 

Mobile Autonomous Robot using the Kinect