基于梯度法的数字散斑图像相关亚像素搜索算法  

Sub-pixel Searching Algorithm of Digital Image/Speckle Correlation Based on Gradient Method

作　　者：王琳霖[1,2] 唐晨[1] 

机构地区：[1]天津大学理学院应用物理系,天津300072 [2]沈阳航空航天大学工程训练中心,辽宁沈阳110136

出　　处：《光电技术应用》2011年第3期45-48,52,共5页Electro-Optic Technology Application

摘　　要：数字散斑图像相关法是对全场位移和应变进行测量的十分有用的无损检测技术。设计精确的亚像素搜索算法是提高测量精度的关键问题。研究了数字散斑图像相关技术的基本原理,在梯度法的基础上,提出了一种基于梯度法的变步长的亚像素搜索算法并应用到相关搜索中,通过对模拟散斑图的位移分析证明了该算法的可靠性和稳定性,并对刚体平移实验进行了测量,结果表明,该算法是稳定可靠、有效的搜索算法,可用于位移场的测量。The digital image/speckle correlation method （DSCM） is a useful nondestructive testing method in the whole displacements and strains. Designing accurate sub-pixel searching algorithms is an important problem in order to enhance the accuracy of measurement. The principle of the digital image/speckle correlation technique is studied, and a kind of gradient method with changeable step is proposed and applied to the correlation searching. The reliability and stability of this algorithm is proved by the displacement analysis of the computer-simulated speckle images, and the translation of the rigid is measured, the results indicate that this algorithm is a stable and reliable, effective searching algorithm, and it can be used in the measurement of displacement fields.

关 键 词：数字散斑图像相关 梯度法 相关搜索 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] O348.12[自动化与计算机技术—计算机科学与技术]