象棋机器人视觉识别算法研究

完成了象棋机器人的视觉识别算法研究,并进行了实验。对棋盘图像预处理,包括对棋盘图像颜色模型的研究、将彩色图像转换成灰度图像、利用均值滤波和中值滤波等方法对棋盘图像滤波、将灰度图像二值化等。利用图像分割算法进行了棋盘图像边缘提取等操作;利用改进的Hough变换对棋子的圆形和圆心进行定位,从而获得棋盘上的棋子位置,为棋子识别打下基础。针对象棋棋子图像特点提出了统计径向棋子像素的识别方法,首先利用数学形态学方法对棋盘图像进行细化处理,然后统计每个棋子的像素个数,从而实现棋子识别。实验结果表明,该方法能得到较好的识别率,而且对于任意角度的棋子像素值均具有保持不变的特点。该方法比较适合象棋机器人视觉系统。

Visual Recognition Algorithm of Chinese Chess Robot

Introduces a chess robot visual recognition algorithm and carries out the experiment.Accomplishes the preprocessing of chessboard,which includes the study of chessboard image color model,transformation from color image to gray image,image filtering with mean filter and median filter,binaryzation of gray image,etc.With image segmentation algorithm to extract the edge of chessboard;With advanced Hough transformation to locate the circle and the center,so obtains the position of chessman on chessboard,makes the base of chessman recognition.Towards the feature of chessman image,proposes the recognition method named radial chessman pixels statistical analysis.The algorithm firstly uses mathematical morphology on chessboard image refining,then get the statis tic of chessman pixels,so achieves the chessman recognition.The experiment results show that the algorithm has good recognition rate,and the pixel value of arbitary angle of chessman keeps un changed.The algorithms is suited for visual recognition of chess robot.