I would appreciate your help for choosing the appropriate real-time SLAM algorithm.

I am interested in having autonomous robot, thus a SLAM component is needed for autonomous navigation.

The platform is Kobuki robot (Turtlebot2 without Kinect) which has wheel odometry. To get a better localization, I want to use camera (Fisheye) - optionally looking upwards.

There is a lot of algorithms available online, most of them output 6DoF (do I need 6DoF for a wheeled robot? (x,y,theta) is enough, right?). I tried using several algorithms, but still didn't find the optimal one.

I believe that Loop-Closure is needed as well, because the final application would be to have a robot that coverage the area (visit each point once).

ROS is used to implement this robot, with small dev board (Intel Atom x5 processor).

Thanks in Advance

  • 2
    $\begingroup$ Can you instrument the environment with fiducials on the ceiling? $\endgroup$ – Ben Feb 11 '17 at 18:02
  • $\begingroup$ Thanks for answering. No, I cannot do that. The purpose is to have robot that can navigate in unknown environment . $\endgroup$ – user3294823 Feb 11 '17 at 19:39
  • $\begingroup$ Can you please clarify what do you mean by "There is a lot of algorithms available online, most of them output 6DoF" ? To be more specific what do you mean by 6DoF output $\endgroup$ – vacky Feb 12 '17 at 18:30
  • $\begingroup$ @vacky That is like (x,y,z, alpha, beta, gamma) unlike (x, y, theta) that OP mentions. 6 DoF planning is generally used for RC planes as far as I know. $\endgroup$ – Prasad Raghavendra Mar 29 '17 at 11:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.