Dynamic movement of multi-agent robot soccer based on local visualization

Dynamic movement is needed for challenging robots which play soccer for obtaining an effective trajectory for individual robot. In this paper, we use exactly 2 agent robots which can be moved using autonomous holonomic movement. Based on visualization, the position of multi-agent robot soccer is localized by image processing from color marker on each of them, It means that their coordinate marker is useful not only for getting their attitude such as angle, both heading angle and angle of target, but also deriving a distance. Furthermore, the data from visualization processing is calculated to be parameter of Vector Distribution Equation which is important for making dynamic movement such as linear movement, angular movement and mixed of them. As a result, multi-agent robot soccer can be precisely moved to the target and has error approximately 1.44% for each trial.