数字图像相关法中散斑图的质量评价

Quality Evaluation of Speckle Image in Digital Image Correlation Method

散斑质量对数字图像相关法中位移的测量精度有着重要的影响,因此定量评价散斑图像的质量是数字图像相关法中一个重要的问题。论文基于灰度共生矩阵提取了散斑图像中的能量、对比度、相关性和熵这四个特征参数,并用实验验证了位移测量结果的均值误差和这四个特征参数之间的关系,证明了这四个特征参数的有效性;同时为了保证散斑质量评价的准确性,论文以平均灰度梯度作为散斑质量评价的基础参数,结合能量、对比度、相关性和熵这四个特征参数,一同作为神经网络的输入参数,经过训练可得到评价散斑图像质量的分类模型。结果表明,该方法可对散斑图像的质量进行较准确的评价。

Speckle quality has an important influence on the accuracy of displacement measurement in digital image correlation method.Therefore,quantitative evaluation of speckle image quality is an important issue in digital image correlation method.In this paper,four characteristic parameters,energy,contrast,correlation and entropy,are extracted from speckle image based on gray level co-occurrence matrix,and the relationship between the mean error of displacement measurement results and these four parameters are verified by experiments.Meanwhile,in order to ensure the accuracy of speckle quality evaluation,the average method is adopted in this paper.Gray gradient is the basic parameter of speckle quality evaluation.It combines energy,contrast,correlation and entropy as the input parameters of the neural network.After training,a classification model for evaluating the quality of speckle image can be obtained.The results show that the method can accurately evaluate the quality of speckle image.