【摘要】 针对直齿轮齿距偏差测量难度大、测量精度低的现状,文章采用机器视觉技术对直齿轮的齿距偏差进行检测。通过机器视觉系统获取齿轮图像,利用改进的Zernike矩亚像素边缘检测算法进行齿轮边缘检测;利用重心法求取齿轮几何中心,利用统计连通域法求取齿数,利用凸包法计算齿顶圆半径并计算获取齿根圆半径,通过齿顶圆公式计算模数,通过分度圆公式计算分度圆半径;依据测量结果给出了齿轮齿距偏差的测量方法,通过测量结果与直齿轮实际尺寸对比和分析,证明了该算法的合理性,可以实现齿距偏差的有效检测。更多还原

【Abstract】 Aiming at the current situation of high difficulty and low accuracy in measuring the spur gear pitch deviation, this paper detects the spur gear pitch deviation based on machine vision technology. The gear image is acquired by the machine vision system, and the improved Zernike moment sub-pixel edge detection algorithm is used to detect the gear edge; The geometric center of the gear is obtained by the center of gravity method, the number of teeth is obtained by the statistical connected region method, the radius of the tooth crest circle is calculated by the convex hull method, and the radius of the tooth root circle is calculated. The modulus is calculated by the formula of the tooth crest circle, and the radius of the indexing circle is calculated by the formula of the indexing circle; According to the measurement results, the measurement method of gear pitch deviation is given. Through the comparison and analysis between the measurement results and the actual size of spur gears, the rationality of the algorithm in this paper is proved, and the effective detection of gear pitch deviation can be realized.更多还原