【摘要】 单目相机定位大尺度矩形工件时,相机畸变严重,图像分辨率高,不适用于实时性和精确度要求高的工业环境,为此,提出一种把工件的对角区域作为研究对象的双目视觉定位方法。首先,根据工件的成像特点,运用改进的Hough直线检测算法得到工件的边缘直线;然后计算工件的角点和旋转角度;最后,将计算结果应用到标定过的双目视觉引导系统中,实现工件的定位抓取。实验结果表明:提出的定位方法快速、准确,抗噪能力强,大尺度工件定位误差小于0.5 mm,重复定位误差小于0.05 mm。更多还原

【Abstract】 When monocular camera locates large-scale rectangular workpiece, the camera distortion is serious and the image resolution is high, which cannot be used in an industrial environment with high real-time and high precision requirements. Therefore, a binocular vision positioning method is proposed, which takes the diagonal region of workpiece as the research object. Firstly, according to the imaging characteristics of the workpiece, the improved Hough transform is used to detect the edge line of the workpiece. Secondly, the corner point and the rotation angle of the workpiece are calculated. Finally, the calculation results are applied to the calibrated binocular vision guidance system to grab workpiece. The results show that this method is fast, accurate and has strong anti noise ability. The positioning error of large-scale workpiece is less than 0.5 mm, and the repeated positioning error is less than 0.05 mm.更多还原