【摘要】 利用机器视觉方法对表面粗糙度的检测进行了研究,提出一种新的环境光补偿法来提高在不同外界环境光情况下的检测精度.利用照度计测量外界环境光的强度变化,并选取工件采样区域特征均值及灰度共生矩阵能量值拟合得到工件表面粗糙度检测值的计算多项式.与算术平均偏差法及灰度共生矩阵法进行了实验对比,对38mm磨削轴在不同环境光下检测结果表明,当测量误差选定为±0.05μm时,环境光补偿法检测结果的正确率比其他2种方法的正确率提高约20%.更多还原

【Abstract】 The non-contact and non-destructive surface roughness inspection method based on machine vision is studied to assess different grinding surfaces in different ambient light environment.A new compensation method is proposed to improve the inspecting results.The light meter is used to measure the intensity of ambient light.Then the intensity of ambient light is fitted with extracted features of the work piece images to calculate corresponding surface roughness.The experiment comparison is made between this new method and other two methods,which are separately based on the arithmetic average absolute deviations and the gray-level co-occurrence gray matrix,to inspect work pieces in different ambient light conditions.Results from an example show that accurate ratio of the proposed method increases about twenty percent compared with the other two when inspecting error of ±0.05 μm is set.更多还原