Research on fast recognizing ball method based on multisensor data fusion in robotic game

In order to solve the puzzle of fast recognizing target ball in the environment of autonomous basketball robot game, the paper proposed a new method of fast recognizing target ball based on multi-sensor data fusion. In the paper, it researched on the criterion of correctly recognizing target ball, on the one hand, it obtained the approximate direction angle α of a single ball by using the visual camera, and on the other hand, it acquired the direction angle β and distance d of some balls by using a 2-D scanning laser range finder in its measured range, if the absolute value of the difference between α and β is within the setting range, the ball is the target ball collectively identified by both the visual camera and 2-D scanning laser range finder. Under the condition of laboratory, the basketball robot game environment has been built, the actual experiment results testified the feasibility and validity of the method. The experimental results show that the proposed method can obtain better effect of correct recognizing ball.