2004-2015 University of Aveiro
- Send Comments to Antonio J. R. Neves
The architecture of our library is of the type ``plug and play'', meaning that it offers support for different vision sensors technologies and the software created using the library is easily exportable and can be shared between different types of vision sensors. UAVision is appropriate for being used by robots that need to detected color coded objects. Another important aspect of our library is that it takes into consideration time constraints. All the algorithms behind this library have been implemented focusing on maintaining the processing time as low as possible. In Robotics, the notion of realtime is an extremely important issue and even though there is not a strict definition of realtime, almost always it refers to the amount of time elapsed between the acquisition of two consecutive frames. Realtime processing means processing the information captured by the vision sensors within the limits of the frame rate.
The vision system for color-coded object detection within the RoboCup soccer games of Middle Size League that we have implemented using the UAVision library can work with frame rates up 50fps using a resolution of $1024 \times 1024$ pixels, both in Bayer, RGB or YUV color modes.
The library that we are proposing comes as a natural development of the work already presented within the RoboCup community. After having implemented vision systems for robotic soccer players that perform both in the Standard Platform League and Middle Size League, we are proposing this new cross-library that can be used by robots whose architecture might be different, but the goal remains the same: detect colored objects.
Alina Trifan, Antonio J. R. Neves, Bernardo Cunha, Jose Luis Azevedo, Real-Time Color Coded Object Detection Using a Modular Computer Vision Library, CSIJ Advances in Computer Science: an International Journal, Vol. 5, Issue 1, No.19, p. 110-123, January 2016.
Antonio J. R. Neves, Alina Trifan, Paulo Dias, José Luís Azevedo. Detection of Aerial Balls in Robotic Soccer Using a Mixture of Color and Depth Information. Proceedings of the 2015 IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC 2015), Vila Real, Portugal, April 2015
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
|
2004-2015 University of Aveiro - Send Comments to Antonio J. R. Neves |