基于色彩均匀敏感度的磨削表面粗糙度测量  

Grinding surface roughness measurement based on color uniformity sensitivity

作　　者：杨晨 方红萍[1] 邹凌云 伍世虔[1] YANG Chen;FANG Hong-ping;ZOU Ling-yun;WU Shi-qian(College of Information Science and Engineering,Wuhan University of Science and Technology,Wuhan 430081,China;Mechanical&Electrical Engineering College,Nanchang Institute of Technology,Nanchang 330099,China)

机构地区：[1]武汉科技大学信息科学与工程学院,湖北武汉430081 [2]南昌工程学院机械与电气工程学院,江西南昌330099

出　　处：《机电工程》2021年第6期755-761,共7页Journal of Mechanical & Electrical Engineering

基　　金：国家自然科学基金资助项目(61775172);江西省教育厅资助项目(GJJ180959);武汉科技大学重大科技培育类资助项目(2018TDX06)。

摘　　要：对于纹理较弱、粗糙度较低的磨削加工表面,采用传统的粗糙度测量方法存在提取表面纹理特征困难,且测量精度受表面纹理方向的影响,对此提出了一种对纹理方向不敏感的彩色光测量方法。首先,基于Beckmann-Spizzichino光反射模型,分析了粗糙面的光反射机理;然后,将磨削加工表面作为反射面,提出了一种基于彩色光反射图像的磨削表面粗糙度测量方案,并搭建了测量平台;最后,结合磨削表面存在的纹理方向特性,设计了一种色彩均匀敏感度测量指标Cu sd,建立了与磨削表面粗糙度的强相关性,实现了对磨削表面的粗糙度测量。实验结果表明:指标Cu sd性能明显优于其他对比指标;与粗糙度的相关性达到0.978,具有较强的正相关性,且该指标对纹理方向不敏感;对粗糙度Ra范围在0.017~0.646μm的磨削表面,指标Cu sd预测均方差为0.00084,平均测量误差为0.02μm,可满足磨削表面测量要求。For the grinding surface with weak texture and low roughness,the traditional machine vision roughness measurement method was difficult to extract surface texture features,and the measurement accuracy was affected by the direction of surface texture,thus a color light measurement method insensitive to the texture direction was proposed.Firstly,based on the Beckmann-Spizzichino light reflection model,the light reflection mechanism of rough surface was analyzed.Then the grinding surface was used as the reflective surface,a plan of color light reflection image-based grinding surface roughness measurement was proposed,and the measurement platform was built.Finally,in combination with the texture direction characteristics of the grinding surface,a color uniformity sensitivity measurement index Cu sd was designed,which established a strong correlation with the grinding surface roughness and realized the measurement of the grinding surface roughness.The experimental results show that the performance of the measurement index Cu sd is better than other gray-scale indexes;the correlation with the roughness is 0.978,which has a strong positive correlation,and the index is not sensitive to the direction of texture;for the grinding surface with the roughness range of 0.017~0.646μm,the mean square error of prediction was 0.00084 and the average measurement error was 0.02μm,which meets the requirements of grinding surface measurement.

关 键 词：磨削表面粗糙度 色彩均匀敏感度 粗糙度测量 光反射模型 纹理方向 

分 类 号：TH741.3[机械工程—光学工程]