Abstract :
We develop the system, which automatically carries out Go-record from the image, which picturized the Go-game situation. Go-stone has to decide its position on the 19/spl times/19 cells of the Go-board to create a Go-record. Consequently, in the first step, only the contour of the Go-board has to be extracted from the image, which includes many unrelated objects. In this paper, we propose the algorithm for the detection of the Go-board contour in the real image using the genetic algorithm (GA). In the proposed algorithm, we show the elite saving strategy mode for combining a global search with a local search. It is the technique, which supplements the defect of a global search. In addition, we show the technique, which executes the proposed algorithm two times for the detection of Go-board contour with more accuracy. We evaluated the proposed algorithm by dealing with the real image, which includes many unrelated objects. As a result, it has been verified that the proposed algorithm detects the Go-board contour very well.
