User:Jdominguez
- Project Name: Localization probabilistic visual in a robot with the simulator Gazebo.
- Authors: Jose Manuel Domínguez Arroyo (joshdomin [at] gmail [point] com) and Jose María Cañas Plaza (jmplaza [at] gsyc [point] es)
- Academic Year: 2008-2009
- Degree: grad
- Jde Version: jde-4.3
- SVN Repository: http://svn.jde.gsyc.es/users/jdominguez/visual_loc_3D
- Technology: c, c++, jde suite, OpenGL, Gazebo,GTK
- Source License: GPLv3
- State: Finished
- Documentation:
Master Thesis: PDF Presentation: PDF
Contents |
History of visual_loc 3D
22-07-2009
03-07-2009
12-06-2009
Error: bad URI in <img>!
Error: bad URI in <img>!
27-05-2009
These pictures show the probability grid algorithm. Each one of the picture show a route of the robot. The first route make a observation in front wastepaper, a turned right 90 grades and a incorporated a movement, a moved of 6400 and a incorporated a movement, a turned right 90 grades and a incorporated a movement, a observation, a turned right 90 grades and a incorporated a movement and a observation. Error: bad URI in <img>!
The second route make a observation since the final of hallway, a moved of 11500 and a incorporated a movement, a observation, a moved of 11500 and a incorporated a movement, a turned right 90 grades and a incorporated a movement, a observation, a moved of 6000 and a incorporated a movement and a observation.
Error: bad URI in <img>!
18-05-2009
These pictures show the probability grid algorithm. The first picture show the robot in front of the wastepaper, and I make a observation. The second picture show the robot turned right 90 grades,I incorporate a movement. The third picture show the robot moved until the end of the hallway, I incorporate a movement. The fourth picture show the robot at the end of the hallway looked at to the other hallway, I incorporate a movement. The fifth picture show the robot at the end of the hallway looked at to the other hallway, I make a observation. The sixth picture show the robot at the end of the hallway looked at to the wastepaper, I incorporate a movement. The seventh picture show the robot at the end of the hallway looked at to the wastepaper, I make a observation
29-04-2009
This video shows the robot in movement with the algorithm of objects.the probability use is alpha*probability_acumulation+(1-alpha)*probability_now
This picture shows the operation of the algorithm of distance in eight angles, 0, 45, 90, 135, 180, 225, 270, 315. This algorithm is the same that the previus but if the result is between 0.7 and 0.1 don't put on to 0.1.
The plots show for one position of the robot all the theta of vision.
In first plot the robot theoretical is in middle of three posts, in the second plot the robot theoretical is in front of the wastepaper and in the last plot is in front of a door.
The first picture show the three plots with the robot real is in middle of the posts.
The second picture show the three plots with the robot real is in front of the wastepaper .
The third picture show the three plots with the robot real is in front of a door.
20-04-2009
This picture show the GUI in GTK for this project.
These pictures show the operation of the algorithms of distance in eight angles, 0, 45, 90, 135, 180, 225, 270, 315.These two algorithms are based in objects instead of pixels.
This picture match with the algorithms of distance make for Alberto Lopez in your project.
This picture match with the algorithms of distance object, it make for me. I get of the image real and theoretic the objects that they aren't objects wall, I compare the width of the objects if they have the equals color, and I get the pixels equals and differents. For I get el weight I do the function equals object X-diferents object X/width object real X,plus all weight/number objects. if the result is minus that 0.7 the distance is 0.1 if not distance is the result.
This video show a teleoperation of the robot for the departamental, too show the observation model with a image theoretic.
20-03-2009
These pictures show the operation of the algorithms of distance in eight angles, 0, 45, 90, 135, 180, 225, 270, 315.
This picture match with the algorithms of distance simple.
This picture match with the algorithms of distance simple smooth.
This picture match with the algorithms of distance without white.
This picture match with the algorithms of distance without white smooth.
This picture match with the algorithms of distance mahalanobis.
This picture match with the algorithms of distance average.
03-03-2009
Table with the six algorithms of distance and the time pass since that it create the first particle, calculating the teorical observation and the distance, until the last particle. The measuring was taking with 6500 particles, vision angle 47, number sample vision 47, sigma radius 500 and sigma theta 15.
Photographs of distance algorithms between a theorical image and real image.
the distance is a % of equals pixels.
the distance is a % of equals pixels, but the theorical image is smooth with the formula (pixelminus2*1+pixelminus1*2+pixel*3+pixelplus1*2+pixelplus2*1)/9.
the distance is a % of pixels no white that they are equals.
the distance is a % of pixels no white that they are equals, but the theorical image is smooth with the formula (pixelminus2*1+pixelminus1*2+pixel*3+pixelplus1*2+pixelplus2*1)/9.
the distance is the distance of mahalanobi between pixels.
the distance is the distance average between pixels
19-02-2009
Photographs of distance algorithms between a theorical image and real image.
the distance is a % of equals pixels.
the distance is a % of equals pixels, but if the theorical image is white the distance is 0.
26-01-2009
How much is discriminate my observation model with filter color? Here I show two examples of my observation model, the first is with the robot view a door and a wastepaper, the other example is view a door.Each example has two images a in gazebo and other in jde.
Error: bad URI in <img>!
Error: bad URI in <img>!
The conclusion is that this observation model isn't discriminate.
17-12-2008
Presentation of a first version of the algoritm visual_loc_3D in jde-4.3.
