This paper proposes a new approach for processing measured data from active Infra Red (IR) thermography, where a soft sensing algorithm is exploited for in depth defect reconstruction. This is achieved by propagating the information gathered at the wall surface to the inner layers. Correlating the experimental 2D measurements to a Finite Element (FE) model of the tested specimen it is possible to update the model with the measured data and change the geometry of the simulated inner defect, until the surface temperature distribution calculated corresponds to the measured one. Following that strategy, the unknown defect geometry can be determined. The method developed and presented in this paper consists of an optimization problem based on the minimization of the difference between the surface temperature distribution measured on the sample subjected to an active thermography test and the one resulting from the FE model. The optimization variables are the geometrical parameters (d...

中文翻译：

---

红外热成像主动数据对深度缺陷几何的软测量重建

本文提出了一种新的方法来处理来自有源红外（IR）热成像的测量数据，其中采用了一种软传感算法来进行深度缺陷重建。这是通过将在壁表面收集的信息传播到内层来实现的。将实验2D测量值与被测样品的有限元（FE）模型相关联，可以用测量数据更新模型并更改模拟内部缺陷的几何形状，直到计算出的表面温度分布与测量值相对应为止。按照该策略，可以确定未知的缺陷几何形状。本文开发和介绍的方法包括一个优化问题，该问题基于最小化了在进行主动热成像测试的样品上测得的表面温度分布与有限元模型产生的表面温度分布之间的差异。优化变量是几何参数（d ...