METHOD FOR AUTOMATICALLY TUNING AND UPDATING ROBOT COLOR FOR ROBOT SOCCER SYSTEM

摘要

PURPOSE: A method for automatically tuning and updating a robot color for a robot soccer system is provided to improve a recognition rate of a robot in the whole area of a playing ground and to reduce a time for a color tuning by automatically adjusting a difference of color data by the different degree of brightness according to sections of the ground and setting a robot color area capable of being recognized in common. CONSTITUTION: If an automatic update program is executed, a robot is recognized through a vision algorithm during a game(S300). RGB data of a robot color are extracted from a color patch of a recognized robot(S310). The mean value is calculated from the RGB data of pixels included in the color patch(S320). The mean value of the RGB data is converted into a YUV form(S330). The converted mean value is compared with an initial robot color area of the YUV form(S340). If the converted mean value is not included in the initial robot color area, a robot color area is updated. Thus, new color area including the mean value is reset by referring to a scope of an initial robot color area(S360). The new color area is compared with the initial color area, only varied amount is stored(S370).