UAVision computer library

This page contains the sequences and software presented in the papers below.

Test sequences

Sequence with the ball stopped and the robot moving in a path (each frame 1024x1024 pixels, RGB)
Sequence with the ball moving on the field and the robot stopped (each frame 1024x1024 pixels, RGB)
Sequence with the ball moving on the field and the robot moving in a path (each frame 1024x1024 pixels, RGB)
Sequence with the ball stopped and the robot moving in a path (each frame 1024x1024 pixels, single channel, Bayer Format)
Sequence with the ball moving on the field and the robot stopped (each frame 1024x1024 pixels, single channel, Bayer Format)
Sequence with the ball moving on the field and the robot moving in a path (each frame 1024x1024 pixels, single channel, Bayer Format)
Binary configuration file with relevant information about the vision system (see configuration file description)

Configuration File Description (for any question contact Alina Trifan (alina.trifan@ua.pt) or Antonio Neves (an@ua.pt)

Related publications

Antonio J. R. Neves, Alina Trifan, Bernardo Cunha. Self-calibration of colormetric parameters in vision systems for autonomous soccer robots RoboCup 2013: Robot Soccer World Cup XVII, Eindhoven, Netherlands, vol. LNCS 8371, p. 183-194, September 2014

Antonio J. R. Neves, Alina Trifan, Bernardo Cunha. UAVision: Modular Time Constrained Vision Library for Color-coded Applications. Proc. of Computational Modeling of Objects Presented in Images: Fundamentals, Methods and Applications, Pittsburgh, PA, USA, vol. LNCS 8641, p. 351-362, September 2014

Alina Trifan, Antonio J. R. Neves, Bernardo Cunha, Jose Luis Azevedo. UAVision: A modular time-constrained vision library for soccer robots, RoboCup 2014: Robot Soccer World Cup XVIII, João Pessoa, Brazil, vol. LNAI, July 2014

Software

The software used in the papers referred above can be found in UAVision source code. To compile the code, enter the build directory and type cmake .. and then type make. There are two applications in the folder app and the corresponding binaries after compilation go to the bin directory.

To use the examples provided on this website (vision and visionCalib applications), the following commands have to be executed:

./vision -cf "configuration-file" -load "video-file-to-load" -debug -server

When in server, a remote client called *visionCalib* can connect to this server with two different purposes:
*remote color calibration
*receive information in real time about what the robot is seeing.

The command to be executed when running visionCalib is the following:
./visionCalib -addr "IP to which one wants to connect"

When running the vision server and visionCalib client on the same host, the IP is 127.0.0.1 .

Other command line parameters that can be used when running vision:
-h (help)
-nodisp (do not show image)
-server (accept connections from imageCalib app)
-cf # (name of the configuration file)
-debug (shows visual debug information on the image)
-v (verbose)
-cam # (camera type 0 - eth; 1 - opencv; 2 - firewire;
-port # (port number when in server mode)
-load # (load a video file)
-save # (save a video file)
-loop (play video file in loop mode)
-fs (full size display)
-slave (vision does not do PMAN_tick so that we can use cursor)
-ball # (ball color configured in visionCalib 1 - orange 2 - yellow 3 - magenta 4 - cyan 5 - blue; if nothing is specified, the orange ball is used by default)

When vision runs in display and debug mode (./vision -debug), the following keys can be used:
1 - shows the original image.
2 - shows the "index" image (grayscale).
3 - shows the color segmented image (RGB).
4 - shows the "reality of the robot" image (RGB).
p - pause image.
f - freeze image.
g - go image (every hit on the g key displays only a frame, after a freeze for example).
a - starts the auto-calibration of the camera settings.
h - displays the help image on the original image.
q - quit.

Important note: For the key interaction to work properly, make sure that the image window is active (to activate a window,click on the image displayed within the window).

Other command line parameters that can be used when running visionCalib:
-h (help)
-addr # (address of the server)
-port # (port number)
-mask # (filename with a mask)

The following key interactions are supported:
i - request an image from the server.
c - request the color ranges from the server.
m - receive the mask from the server.
s - receive the camera settings from the server.
y - the trackbars and HSV histogram pop up.
C - send the color ranges to the server, where they will be saved on the configuration file.
M - send the mask to the server, where it will be saved on the configuration file.
S - send the camera settings to the server, where they will be saved on the configuration file.
a - start the auto-calibration of the camera settings on the server side.
e - manual calibration of the camera settings using sliders.
h - opens the help menu in a different window.
q - quit.

Important note: before starting any color calibration, make sure you have requested the image, color ranges and mask from the server.