数字图像相关法散斑图质量评价方法  

Assessment method of speckle pattern quality in digital image correlation

作　　者：俞海 刘云鹏[2] 郭荣鑫[2] 夏海廷[2] 颜峰[2] YU Hai;LIU Yunpeng;GUO Rongxin;XIA Haiting;YAN Feng(School of Civil Engineering,North Minzu University,Yinchuan 750021,China;Department of Engineering Mechanics,Kunming University of Science and Technology,Kunming 650093,China)

机构地区：[1]北方民族大学土木工程学院,银川750021 [2]昆明理工大学工程力学系,昆明650500

出　　处：《激光技术》2020年第2期237-243,共7页Laser Technology

基　　金：宁夏高等学校科学研究资助项目(NGY2017153);宁夏自然科学基金资助项目(NZ17116)。

摘　　要：为了研究数字图像相关方法中散斑图质量的有效评价问题,采用散斑图质量表征参量——平均灰度2阶导数对散斑图质量进行了有效评价。通过分析图像插值误差与图像灰度信息分布形式之间关系,指出散斑图平均灰度2阶导数与散斑图灰度信息分布形式之间的关系;为了验证该散斑图质量表征参量的有效性,利用傅里叶变换对散斑图进行平移实验,通过Newton-Raphson(N-R)方法对平移前后的散斑图进行亚像素位移计算,由位移计算结果可知,具有低的平均灰度2阶导数的散斑图对应小的位移测量误差。结果表明,平均灰度2阶导数在散斑图质量评价中具有一定的有效性;在实际应用中,应结合散斑图平均灰度2阶导数和平均灰度梯度对散斑图质量进行综合有效的评价。该研究为高质量散斑图的制备与选取提供了参考。In order to study the effective evaluation of speckle image quality in digital image correlation method,the quality of speckle pattern was effectively evaluated by means of the second derivative of average gray level of quality characterization parameter of speckle pattern.By analyzing the relationship between image interpolation error and the distribution of image gray information,the relationship between the second derivative of average gray level of speckle pattern and the distribution form of gray level information of speckle pattern was pointed out.In order to verify the validity of the quality characterization parameters of the speckle pattern,Fourier transform was used to translate the speckle pattern.The sub-pixel displacement of speckle pattern before and after translation was calculated by Newton-Raphson method.According to the displacement calculation results,speckle pattern with low average gray second derivative corresponded to small displacement measurement error.The results show that,the second derivative of average gray level is effective in the quality evaluation of speckle pattern.In practical application,the quality of speckle pattern should be evaluated comprehensively and effectively by combining the second derivative of average gray level and average gray gradient of speckle pattern.This study provides a reference for the preparation and selection of high quality speckle patterns.

关 键 词：测量与计量 数字图像相关 插值误差 散斑图质量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN247[自动化与计算机技术—计算机科学与技术]