JDERobot

In the next video you can see running JDE 4.3 with usb camera in pico-itx.

Today we’ve watched the “Revolution OS”, a 2001 documentary. It traces the history of GNU, Linux, and the open source and free software movements. It features several interviews with prominent hackers and entrepreneurs (and hackers-cum-entrepreneurs), including Richard Stallman, Michael Tiemann, Linus Torvalds, Larry Augustin, Eric S. Raymond, Bruce Perens, Frank Hecker and Brian Behlendorf.

The film begins in medias res with an IPO, and then sets the historical stage by showing the beginnings of software development back in the day when software was shared on paper tape for the price of the paper itself. It then segues to Bill Gates’s Open Letter to Hobbyists in which he asks Computer Hobbyists to not share, but to buy software. Richard Stallman then explains how and why he left the MIT Lab for Artificial Intelligence in order to devote his life to the development of free software, as well as how he started with the GNU project. Linus Torvalds is interviewed on his development of the Linux kernel as well as on the GNU/Linux naming controversy and Linux’s further evolution, including its commercialization. Richard Stallman remarks on some of the ideological aspects of open source vis-á-vis Communism and capitalism and well as on several aspects of the development of GNU/Linux. Michael Tiemann (interviewed in a desert) tells how he met Stallman and got an early version of Stallman’s GCC and founded Cygnus Solutions. Larry Augustin tells how he combined the resulting GNU software and a normal PC to create a UNIX-like Workstation which cost one third the price of a workstation by Sun Microsystems even though it was three times as powerful. His narrative includes his early dealings with venture capitalists, the eventual capitalization and commodification of Linux for his own company, VA Linux, and ends with its IPO.

You can see it on: http://video.google.es/videoplay?docid=7707585592627775409 (this video is available under the terms of the GNU Free Documentation License )

Pico-Itx

This board is 100% compatible with linux (debian, ubuntu). We have managed to install Debian Sid in this board without problems. Moreover we have installed the new version of JDE (4.3) by means of debian packages and everything has worked well.

In the next screenshot you can see a jde execution in pico-itx motherboard:

running jde4.3 in picoitx

We have a mini-itx mother-board and touch tft 7″ too. We will describe the experiences with this gadgets in the nexts weeks.

Title: The CarSpeed project
Speaker: Victor Hidalgo
Date: 2008/09/30, 19h
Place: Universidad Rey Juan Carlos at Robotics Lab (Departamental-II)

Next tuesday, September 30 at 19h, we will hold the seminar “The CarSpeed project’ in the Robotics Lab. Victor Hidalgo will explain his work in that project, and the underlying computer vision technology. CarSpeed deals with an application that it is able to carry out precise estimation of vehicle speeds. It also counts the number of the vehicles that travel through the road.

The system uses a single off-the-shelf camera. Its images are carefully analyzed to give estimations of vehicles speed. The application detects and tracks the vehicles on the road. The detection step allows the system to realize the appearance of new vehicles. It uses a sampled motion filter. The tracking algorithm starts for each new detected vehicle. It implements an homography and an evolutionary algorithm that keep several speed hypothesis while they are compatible with the camera images. Check out this video (right) for a demo.

I’ve finished my project. Our approach specifically addresses issues such as safe navigation in unmodified and dynamic environments, like Departamental II of this university. We’ve solved the following problems:

- Navigation in dynamic environments. Public places are often packed with people. People behave not necessarily cooperatively. Our approach provides means for safe and effective navigation through crowds.

- Navigation in unmodified environments. No modification of the environment is necessary for the robot’s operation.

- Localization. In every operation, our robot continuously tracks its position using its maps. Position estimates are necessary for the robot to know where to move when navigating to a specific goal, and to ensure the robot does not accidentally leave its operational area.

To navigate reliably in indoor environments, a mobile robot must know where it is. Thus, reliable position estimation is a key problem in mobile robotics. We believe that probabilistic approaches are among the most promising candidates to providing a comprehensive and real-time solution to the robot localization problem.

So, in this video we’ve used Monte Carlo localization method where we represent the probability density involved by maintaining a set of samples that are randomly drawn from it. We show experimentally that the resulting method is able to efficiently localize a mobile robot without knowledge of its starting location.

Link for more information: http://jde.gsyc.es/index.php/jmvega_guide_robot

Next Monday, September 22 to 18h, we will hold the seminar “Software Tools in the context of JDE” in the robotics lab.

We explain the motivations of this free software project and tools with which we work, both in JDE in the PFC: svn, mediawiki, blog, tracs, etc.. Explain the uses basics, “good practice” and solve the questions. This seminar is designed for all students who are starting their PFC.

Title: Software Tools in the context of JDE
Speakers: Jose María Cañas and Roberto Calvo
Date: 22 September 2008, 18h
Place: Universidad Rey Juan Carlos in robotics lab (departamental 2)

Our ElderCare application was explained at TeleMadrid news on 2008-07-08.

Eldercare autonomously detects people falls using regular cameras. It is aimed at elder people which live alone. The system reduces the time without medical assistance in the case of fall or faint of the elder patient. It continuously analyzes the images from several cameras and estimates the 3D position of the people in the monitored area. In the event of a fall or a faint it triggers the alarm to alert a relative or a doctor. The system does not require any actuation from the patient, it is completely autonomous and non intrusive. It is based on particle filters technology, motion and color analysis.

Antonio Pineda and Sara Marugán share the credits of this application

More info at this webpage.

This is one sample exercise that alumni from Robotica course did on the second semester of 2008. A pioneer robot in Gazebo simulator had to follow a red line at the floor, using only its own (simulated) camera. They programmed a reactive vision-based controller to do so.

The Skeleton Visualizer has been updated.

Now, the user can pause and replay the movement action of the skeleton (Ramón) by the two buttons directly inserted in the virtual scene. This improvement has been done through the Progeo library, that allow to relate the 2D points corresponding to the imagen pixels with the 3D points of the virtual stage. If the line projected from the pixel go through one of the “virtual” buttons, the action is paused or played.

yqURgIW3_SU

This is a video example of the “skeleton in movement” visualizer that we’re developing. This scheme allows the camera control by mouse and a series of illumination options. Enjoy it.