JDERobot » Uncategorized

Distributed Video Surveillance System based on Android

rocapal — Fri, 23 Jul 2010 14:13:55 +0000

Security and Surveillance Systems have always been demand in the industrial and home environment. The cameras, which human eyes are able to collect all visual information from the environment. Extensions necessary and highly desirable in these systems, and that few possess, are integration and interconnection with other systems, video surveillance and display all the information on a mobile device, which lately is becoming essential in a peripheral daily life. Most existing surveillance systems are based on the hardware, and software is old and have not a good design, so it’s difficult allowing scalability and increased functionality.

This project, called Surveillance, raises the development of a distributed video system integrated in mobile device based on Android. For this we have designed a distributed system, multi-platform, multi-lenguage and high scalability. The system performs records through the cameras installed and it is associated with an alarm generation by motion detection. All this is managed and viewed from a mobile device based on Android with intelligent touchscreen. One of the most important features has been developed is receiving real-time video in mobile device, both the cameras and recordings made.

The system has been experimentally validated and proven correct operation. For the implementation of this project is used software platform architecture and JDEROBOT, distributed component programming six main functions that are: get the images from the camera, motion analysis, centralize the management of the recordings, making the recordings, display information on mobile and manage logical recordings and alarms. It has also opted to use frameworks, consolidated libraries and tools (ICE, GStreamer and VLC) to provide robustness to the system. Finally, we have integrated and improved several components of this infrastructure, obtaining a correct and efficient functioning.

Of sourse, this project is a FLOSS (Free Libre Open Source Software) project. You can obtain the source code and documentation in the next links:

JDEROBOT Card: https://jderobot.org/index.php/Rocapal_Enabling_Technologies_LS
JDEROBOT Sources: https://svn.jderobot.org/jderobot/trunk/
Surveillance App Android: https://svn.jderobot.org/users/rocapal/master/src/SecurityApp/
Master Thesis: [PDF] [Sources]
Slides Thesis: [PDF] [Sources]

Posted via email from rocapal’s posterous

Talk: vision on mobile robots

Julio Vega — Fri, 11 Dec 2009 19:03:06 +0000

Dr. Pilar Bachiller from Universidad de Extremadura has presented the system developed on her PhD in order to get the robot behavior based on visual attention for mobile robots. The design of a control architecture for a mobile robot involves seeking solutions to issues such as analysis and interpretation of sensory information provided about the environment, the limited scope of the sensors, the limited time to process image and maintenance of multiple behavioral objectives.
You can download the PDF slides and also you can see or download full recorded speech on .avi video here.

Seminar: how to compile, link and debug our C applications on linux

Julio Vega — Thu, 21 May 2009 18:50:20 +0000

Today, Jose María Cañas and Julio Vega have given a talk about how to compile, link and debug C/C++ applications over Linux platform.
Maybe sometimes there are many problems about how to compile our applications with gcc (GNU C Compiler) or link them with dynamic libraries and how to check the correct linked. Moreover, we usually prefer to include all this process in a simple makefile. So they’ve explained how to build a makefile.
On the other hand, when we have perfectly build our application or executable, we want to know why our application doesn’t work fine… so we need to debug it. They’ve talked about two important tools: gdb (GNU Debugger) and valgrind. Gdb in order to check the internal structure of the program and valgrind, specially indicated to help us to check dynamic memory use.

You can download the PDF slides and also you can see or download full recorded speech on .avi video here.

jderobot-4.3.0 launched

jmplaza — Thu, 16 Apr 2009 08:38:45 +0000

The new stable release of jderobot has been launched, it is the 4.3.0 release. You can find it here for download or simply install it from debian packages. Many bugs have been fixed, it includes many improvements and new features:

Project code managed with AutoTools
Debian packages created (for ubuntu-hardy, debian-lenny and debian-sid)
Application name changed to JDEROBOT
All the code is now licensed by GPL v3.0
Improved shell
New drivers: video4linux2, evi (for Sony VID30PTZ) and wiimote.
New schemas: a camera calibrator, extrinsics, opencvdemo, wiioperator, etc.

Check its release notes for further details.

In addition, several development tools are now used: mediawiki web site, svn repository, trac…. A community has been gathered around the project. We mainly use two mailing lists for communication: jde-users@gsyc.es and jde-developers@gsyc.es.

Most of the developer team gathered at a pizza party on 2009/03/20 to celebrate the jderobot-4.3.0 launch.

Enjoy!

Seminar: Projective Geometry for computer vision (2009/03/6,17h)

jmplaza — Tue, 03 Mar 2009 18:07:50 +0000

In this talk José María Cañas will present the foundations of projective geometry for dealing with images:
(a) 2D projective geometry, points, lines, homogeneous coordinates.
(b) Image formation: from 3D reality to 2D images, projection matrix, 3D translations and rotations
(c) I will talk about Progeo, a C library used in many JDE applications to relate 2D image pixels with 3D space
(d) Camera calibration, different types of cameras.

Date: 2009/03/06, 17h
Place: Robotics Lab (room 117), Departamental-II, URJC, Móstoles

You can see or download full recorded speech on .avi video here

Finished Skeleton Visualizer Project

davader — Mon, 23 Feb 2009 17:54:08 +0000

I’ve finished my Final Year Project, entitled “Visualizador 3D interactivo para laboratorios de análisis de marcha”. Our objectives were to develop a 3D interactive tool that could help the doctor in the treatments and analysis of patients with walking pathologies.

Concretely, I have programmed a 3D visualizer that shows skeleton figures with the movement captured previously on the real patient. My application reads the file with the position of the patient’s joints and it draw the moving skeleton on a virtual environment with this data.

I have made the application more ergonomic and interactive: there are Play/Pause buttons, a speed dial, a progress bar and some camera controls by the only use of the mouse. this way, the managing is more intuitive and ergonomic.

Finally, the application allows to show some graphics with the temporal evolution of the positions or primary rotation angles of any joint. You can select the choised joint by the upper menu or simply clicking in the joint inside the virtual world.

In this video, you can see the Skeleton Visualizer in all his brilliance:

Last Thursday, I defended my project with great success. I obtained a 10. Thanks to all.

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

Seminar: visual multitarget 3D tracking using particle filters (2009/02/16,17h)

jmplaza — Sun, 01 Feb 2009 11:00:50 +0000

Dr. Marta Marrón from Universidad de Alcalá will present us the algorithms developed on her PhD thesis to visually track several objects using particle filters. The visual 3D tracking of several objects at the same time has many applications in robotics and computer vision. The main advantage of her MultiTargetTracking (MTT) particle filters is that they are truly multimodal, overcoming the only-one-target feature of the classic particle filters. The algorithm she proposes clusters the particle set in different groups according to the several relevant objects in the scene and tracks them with such multimodal single particle population.

Date: 2009/02/16, 17h
Place: Conference Room, Departamental-II, URJC, Móstoles

Visiting the Motion Capture Systems from Vicon (URJC, Alcorcón)

Julio Vega — Fri, 16 Jan 2009 08:55:07 +0000

Yesterday we went to the URJC (Alcorcón) in order to see the motion capture system installed there; and a house prepared to track a person who lives inside the building.

Vicon systems have been doing exactly to track and analyze movement. So, in this hospital they use it to examine the development of patients who have some anomaly on gait. Furthermore, because of Vicon systems is so complex to use it, they have another motion capture system called “Step 32″.

On the other hand, we went to a special house installed there too. In this house, we can be tracked throughout daily life. And then, they get statistical graphics for knowing how we move inside home.

Seminar: Robotic Soccer, from Aibos to Nao humanoids (2008/12/15, 17h)

jmplaza — Tue, 23 Dec 2008 16:39:37 +0000

Some days ago Dr. Humberto Martínez, from Universidad de Murcia, gave us a talk about his experience in the TeamChaos team. TeamChaos team has competed at the RoboCup in the last years, first in the Sony Legged League with Aibos and now in the Standard Platform League, with the Nao humanoid. RoboCup is a very demanding environment for robotics, as long as it is very dynamic, competitive and the onboard computers have limited computing power.

He described the architecture of the software controlling the robot, named ThinkingCap, some locomotion issues and perception algorithms. Locomotion has increased its complexity from the four legged Aibos to current biped humanoids. Speed and stability are now to key topics. Perception is mainly based in vision and must be very efficient in order to allow responsive behaviors.

You can download the slides used in this talk. And you can see or download full recorded speech on .mpg video here

Pioneer wanders around department hall

Julio Vega — Wed, 17 Dec 2008 18:19:49 +0000

About two weeks ago, we recorded Pioneer robot wandering over this department hall. The goal is quite simple: the robot must go from one side to another one. It uses a scanning laser in order to avoid dynamic (e.g. people) and static obstacles (e.g. doors, walls).

Experiments show that algorithm is extremely efficient. We tested the system on a real environment, and it was able to wander around obstacles and people who are walking over there.