机器视觉在激光干涉测量d31中的应用

提出一种应用机器视觉技术的激光干涉法, 可以研究压电材料的逆压电系数. 在传统干涉测量法中引入机器视觉实现干涉图像灰度最佳估计, 使用数字图像处理技术, 一方面应用降噪算法有效去除激光干涉图样中的噪声, 从而确定激光干涉场的光强分布, 另一方面将干涉图样的明暗变化进一步转化为可量度的灰度变化以提高测量分辨力. 通过合理设计算法, 该方法中对位移的理论测量分辨力可以提高达一个数量级, 为干涉光波长的1/1024, 因此在逆压电系数d₃₁ 的实验测量中可以有效提高分辨力.关键词：机器视觉逆压电系数双光束干涉数字图像处理

Application of machine vision to the measurement of the effective d31 coefficient in laser interferometry

A novel method based on machine vision is proposed for measuring the effective d₃₁ coefficient. Based on the best estimation of gray scale of interference pattern, the sensitivity of the twin-beam interferometry is significantly improved in comparison with previously reported system. Firstly, the interference pattern grabbed by machine vision system is extracted using the algorithm developed in this work. Secondly, using gray-scale pattern with 256 gray-levels the measurement accuracy of the interferometric optical path difference of the interferometer can reach 1/1024 of the wavelength of the laser theoretically. So the experimental results of a PZT pipe in the method are greatly accurate.
Keywords：machine visioninverse piezoelectric coefficientdouble-beam interferometrydigital image processing