| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
Hi all I'm on this localization project and I'm required to do localization using a kinect camera. I did some research and all the resources I found are using laser scan instead of a camera, like the ACML package. So I'm wondering if anyone knows a good resource on camera_based localization. Lots of thanks !!!
I did also see a technique somewhere (cannot tell where any more unfortunately) where they just took a slice of their kinect data to get the same data as a laser scanner. So that you can use the kinect as if it where a laser scanner.
I think the package your thinking of is pointcloud_to_laserscan from the turtlebot stack. http://www.ros.org/wiki/pointcloud_to_laserscan
you may have to use VSLAM which refer to Visual Simultaneous Localization and Mapping but this require feedback from vision system and odometer system
most simplified one using only Vision system is by using Bearing Only SLAM
Asked: 2012-08-06 10:43:57 -0500
Seen: 905 times
Last updated: Nov 24 '12
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
