Research On The Game Platform Of Gomoku Man-machine Based On DOBOT Manipulator

With the continuous development of science and technology and the increase in market demand,robots have transformed from traditional industries to ubiquitous consumer-grade products,penetrating into our lives,such as man-machine game.IBM's "Dark Blue" and Google's "Alpha-Go" defeated the then chess world champion Kasparov and No.1 world Go champion Ke Jie,respectively.Man-machine game represents the advanced level of artificial intelligence from a certain aspect.Based on this consideration,a DOBOT Magician robotic arm is used as a carrier to establish a man-machine game platform for backgammon that can be used for research and actual combat.It is meaningful to build a man-machine game platform that can be used for research and actual combat.The platform mainly includes four aspects of Gomoku algorithm,camera calibration,image processing and robotic arm motion control.Based on UCT-based Monte Carlo tree search and combining the features of convolutional neural network in pattern recognition,this paper builds an efficient and stable Gomoku strategy system.Then use Zhang's calibration method to calibrate the camera,reduce the distortion coefficient of the camera,improve the imaging quality,use the Huff transform to extract the chessboard information contained in the image,and combine it with the secondary development of the DOBOT robotic arm to use it as the computer's execution mechanism The robotic arm executes picks and moves until the end of the game.This game platform can be effectively executed and has a high level of chess power.It is an intelligent man-machine game platform,which not only fits the research direction of artificial intelligence in the contemporary era,but also can be used as a puzzle product to bring people the joy of game.