Home /Research /Computer vision system for recognition and detection of color patterns in real-time task of robot control
PERCEPTION

Computer vision system for recognition and detection of color patterns in real-time task of robot control

В. А. Антипов, V. A. Kokovkina, Vasilii Kirnos, Andrey Priorov

Year
2017
Citations
5

Abstract

This work considers the algorithm of mobile robot recognition and localization on the basis of color patterns, applied in robosoccer. For exact position definition of mobile robots and a ball in robosoccer it is necessary to analyze the image received from camera. Whereas each of these objects on the image has the color pattern consisting of circles, the first step of algorithm is detecting of circles. Then circles are classified using the k-means clustering that allows the algorithm to adapt in the conditions of dynamic lighting. Further on the basis of the received circles labels are defined. Also in this work researches of influence of level of illumination and noise on algorithm functioning have been conducted.

Keywords

Computer scienceComputer visionArtificial intelligenceTask (project management)Robot visionRobotMachine visionMobile robotEngineering

Related papers

Browse all PERCEPTION papers