Planning and Control of Humanoid Robots

Vision-Based Odometric Localization for Humanoids using a Kinematic EKF
(software release)

This page contains the software implementation for the vision-based localization algorithm described in [1], and the relative control part described in [2].

The software has been designed to run with the NAO robot from Aldebaran, thus it uses its APIs to access sensor readings. Moreover we command the robot by passing reference values to the Aldebaran built-in walking engine. Adapting the software to any different robot should require a limited effort, provided it is equipped with a sensor suite similar to the one described in [1, 2]. We will be happy to give support and details about our software if needed. If you plan to use this software in a scientific publication, please cite the papers listed at the bottom of the page.

Available software
(Last update: 19 June 2015)

In the zip files listed below you can find instructions and some bash scripts which should help in the installation process. Moreover, we are still working to improve and simplify the installation procedure as much as possible. If you experience any problem, or you have any suggestion, please do not hesistate to contact us.

• HuViO Version 0.1 (19 June 2015)
  
  NOTES
  We developed the version v0.1 under Ubuntu 10.04 (32bit). Anyway, the whole framework is designed to be platform independent. Unfortunately, we still can not ensure a stable interconnection with the Aldebaran or PTAM software. Indeed, we have experienced some problems in compiling the PTAM software under Ubuntu 14.04 64bit. Since PTAM is currently used to run our localization framework we suggest to get it working independently before trying to integrate it in our software. Given the above mentioned issues, we suggest using the 32bit version of Ubuntu. As explained in [1], the use of PTAM is not mandatory, and with some modifications in the software it can be substituted with any other Visual-based Simultaneous Localization and Mapping (VSLAM) or Visual Odometry (VO) algorithm.

[1] G. Oriolo, A. Paolillo, L. Rosa, M. Vendittelli, Vision-Based Odometric Localization for Humanoids using a Kinematic EKF, 2012 IEEE-RAS Int. Conf. on Humanoid Robots, Osaka, Japan, Nov-Dec 2012 (pdf).

[2] G. Oriolo, A. Paolillo, L. Rosa, M. Vendittelli, Vision-Based Trajectory Control for Humanoid Navigation, 2013 IEEE-RAS Int. Conf. on Humanoid Robots, Atlanta, GA, Oct 2013 (pdf).