基于一阶及二阶灰度梯度的散斑图质量评价方法

散斑图质量对数字图像相关(DIC)方法的测量精度具有重要的影响。以平均灰度梯度这一传统的散斑图质量评价参数作为基础,提出了一种新的评价参数——平均一阶及二阶灰度梯度商,该参数同时考虑了一阶和二阶灰度梯度。从整体上来说,一阶灰度梯度的曲面比较光滑,而一阶及二阶灰度梯度商的曲面比较粗糙。采用Zhou等人提出的方法制作了80张模拟散斑图,对这些模拟散斑图的传统参数和提出参数进行了计算。结果表明:当散斑半径相同时,随着散斑数量的增加,两种参数均先增加,后减小;当散斑数量一定时,二者均有一个峰值;散斑半径越大,提出参数的峰值越大,而传统参数的峰值越小。对提出参数的峰值所对应的4张散斑图进行了多次连续平动,并使用DIC方法测量了测点的水平位移。结果表明,在提出参数越大的散斑图上测点的水平位移误差越小,在一定程度上验证了提出参数的正确性。

Method for speckle pattern quality assessment based on one-order and two-order intensity gradients

Speckle pattern quality has an important influence on the measurement accuracy of the digital image correlation (DIC) method.Based on a traditional parameter called the mean intensity gradient,a novel parameter,called the mean of the quotient of one-order and two-order intensity gradients,is proposed for quality assessment of speckle patterns.Both the one-order and two-order intensity gradients are considered in the proposed parameter.Totally,the surface of the one-order intensity gradient is fairly smooth,while the surface of the quotient of one-order and two-order intensity gradients is rough.Eighty simulated speckle patterns are produced by a method proposed by Zhou et al.And the traditional and proposed parameters of the simulated speckle patterns are computed.When the speckle radius is constant,the two parameters first increase and then decrease with an increase of the speckle number,and when the speckle numbers reach certain values,the two parameters reach their peaks.The larger the speckle radius is,the higher the peak of the proposed parameter is,while the lower the peak of the traditional parameter is.Four speckle patterns corresponding to the peaks of the proposed parameters are translated many times,and horizontal displacements of points on the speckle patterns are measured by the DIC method.The higher the proposed parameter is,the lower the bias error of measured displacements of points is.Thus,the proposed parameter is verified to some extent.