UChile1 Strikes Back, 2006 Team Description Paper

Abstract

The UChilel four-legged team is an effort of the Department of Electrical Engineering of the University of Chile in order to foster research in robotics at high level. This document describes the relevant aspects of the UChilel software, which has been developed from scratch and is constantly being updated by our team. The system has shown to be a relatively successful approach; we had an acceptable participation in RoboCup 2003RoboCup and 2006. This year we have improved several aspects of our UChilel software (mainly localization, strategy and actuation). We have also proposed an automated refereeing and analysis tool of robot games. All these aspects are here reported.Index Terms-I. SYSTEM ARCHITECTURE Our software system is divided into four task-oriented modules: vision, localization, strategy and motion control (see figure 1). The vision and motion control modules operate in each robot locally. The localization module is distributed, it operates in each robot, and a global estimate of the overall ball localization is generated in a distributed fashion. The strategy module is also distributed, and allows the sharing of global information among the robots. In our current implementation these four task-oriented modules run, in each robot, in two different computer processes: one for vision and localization that we call the perceptual process Manuscript module, and one for motion control and strategy that we call the engine process module. The reason for having just two processes and not four is the synchronization problems we had in our previous implementation. In the next sections each of the mentioned modules are described. IIVISION MODULE A.Beacons, Goals and Ball PerceptionLandmarks (beacons and goals) and ball are perceived using a color based vision method (most of the RoboCup fourlegged teams use similar vision systems). Robot relative distances and angles are estimated using a segmented image that is built using a look up table and the a priori knowledge of the field objects colors. Our color-based vision system is detailed described in D and in our technical report 2005 D. Fig. 1. Modular organization of our system. In the bottom the low-level processes of vision and motion control. On top the high level processes of localization and strategy. B. Visual Sonar and Lines PerceptionUsing the idea of the visual sonar, images can be analyzed very quickly using scan virtual rays (see figure 2); lines perpendicular to these scan rays can be easily found. Thus, in order to detect a pixel that lies on a line, we check the difference in the Y channel of adjacent pixels in the scan rays. The Y channel describes the luminance of the pixels in a YUV image; strong variations of Y channel indicate a transition from any dark color to white. The image is crossed using a grid of lines perpendiculars to the image horizon. The horizon of the image is obtained by compensating the camera rotation and inclination using the measures of the robot encoders (head 200

Motion Detection and Object Tracking for an AIBO Robot Soccer Player

“…For the experiments, an AIBO robot using the motion software of the UChile Kiltros AIBO soccer team (Ruiz-del-Solar, 2007), configured for allowing just head movements was used. The algorithm runs in the robot in real-time.…”

Task-Oriented Probabilistic Active Vision

“…Of course, this module is very dependent on the robot platform. Our actuation 25 for humanoid robots (tested on Aldebaran Robotics NAO and Hajime HR-18) uses omnidirectional gaits provided by the robot manufacturers.…”

Spatiotemporal Context Integration in Robot Vision

“…The rules verify the geometric and color relationships between the blobs that compose a given object's perception. 19,20 In the case of lines, we need to consider some special features: (i) lines can be detected but not identified (context is required to accomplish this) and (ii) line's position cannot be exactly determined, because only a segment of the line is normally observed. In our case, in order to detect lines, we first detect points of high gradient intensity on the image using scan lines.…”