剪切照相
剪切(物理)
小波变换
小波
光学
曲率
高斯分布
材料科学
相(物质)
干涉测量
数学
声学
人工智能
几何学
计算机科学
物理
复合材料
量子力学
作者
Fabrice R. Michel,Vincent Moreau,Vanessa Rosso,Serge L. Habraken,Bernard Tilkens
摘要
Shearography is a recognized interferometric technique in non-destructive testing to detect defects. Defects are detectable in wrapped phase maps because they are characterized in their neighborhood by singular fringes. They are detectable in unwrapped phase maps, because they induce unexpected phase values. By analyzing the length of unexpected phase values area in shearing direction, and by taking into consideration shearing amount, defect size can be locally estimated. To examine this length, we propose to locally determine borders of unexpected phase values region by analyzing wavelet transform of unwrapped phase map profiles. The borders of defect area are found by examining the convergence at fine scales of lines of wavelet modulus maxima. To have a physical interpretation of this convergence, second derivate of a Gaussian is employed as mother wavelet: estimated borders of defect region are some maximal curvature points of unwrapped phase map profile. To finish, we show that shearing amount does not affect estimated defect size with our methodology. So, shearography is adapted to quantify defects in shearing direction. Currently, in any other direction, an ambiguity exists on the position where the local estimation of defect width is performed. The methodoly cannot be employed.
科研通智能强力驱动
Strongly Powered by AbleSci AI