摘要
棋子定位是下棋机器人视觉系统中至关重要的环节。目前下棋机器人视觉系统中最常用的圆检测算法为霍夫圆变换法。该检测算法具有同时可一次检测多个圆,满足对棋局实时检测的需求等优点,但在暗光照环境下也暴露出不稳定性,导致采集到的图像存在较大的局部阴影和局部反光现象,使得检测出现偏差。为了克服其检测的光敏性,增强检测的鲁棒性,设计了一种图像亮度均衡算法。为了验证方案的可行性,对在暗光源环境下加入图像亮度均衡算法与不加图像亮度均衡算法分别进行了霍夫圆检测的对照实验。实验结果表明,加入图像亮度均衡算法对暗光源环境下采集的棋盘图像预处理后,再进行检测,能够矫正偏差,准确检测到图像中棋子的内轮廓圆,准确实现棋子精准定位。
The positioning of chess pieces is a crucial link in the vision system of chess robot.Hough circle transform is the most commonly used circle detection algorithm in the vision system of chess robot.The detection algorithm has the advantages of detecting multiple circles at the same time to meet the needs of real-time detection of chess games,but it also exposes instability in dark lighting environment,resulting in large local shadows and local reflections in the acquired images,which makes the detection deviation.In order to overcome the photosensitivity and enhance the robustness of detection,an image brightness equalization algo-rithm is designed.In order to verify the feasibility of the scheme,a control experiment of Hough circle detection with and without image brightness equalization algorithm is carried out respectively in the dark light source environment.The experimental results show that after adding the image brightness equalization algorithm to the chessboard image collected under the dark light environ-ment,the error can be corrected,the inner contour circle of the chess pieces in the image can be accurately detected,and the pre-cise positioning of the chess pieces can be accurately realized.
作者
任建松
来枫璟
Ren Jiansong;Lai Fengjing(School of Electromechanical and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074,China)
出处
《现代计算机》
2024年第6期56-60,共5页
Modern Computer
关键词
亮度均衡
霍夫圆变换
下棋机器人
双三次插值
brightness equalization
Hough circle transformation
chess robot
bicubic interpolation