PSI - Issue 20

R.S. Akhmetkhanov / Procedia Structural Integrity 20 (2019) 218–221 R. S. Akhmetkhanov / Structural Integrity Procedia 00 (2019) 000 – 000

220

3

Fig 1. Thermogram image (a); histogram of density of distributions in temperature in thermogram (b); structural image of thermogram (c); and selected area of the field with defect (d).

Fig.2. The isolines of thermogram in the region with an elevated temperature of the sample (a); the boundaries of regions with different temperatures (b); the carcass of the image of the thermogram (c).

Thus, having highlighted the area of defects in the image of the thermogram, we lost information about other areas of the temperature field with small levels of local temperature changes, which also represent local material inhomogeneities. Therefore, in order to identify inhomogeneities in the structure of the material, which are characterized by small fluctuations of the thermal field, a multiple-scale wavelet decomposition of the resulting thermal field into components was applied. If prior to the decomposition we had a standard deviation in the color of the image pixels StdDev = 9.761, then the image components have significant increases in standard deviation. They are estimated at between 12.6 and 38.6.