Neural networks and genetic algorithms for the evaluation of coatings thicknesses in thermal barriers by infrared thermography data

Maximizing the detection of thermal imprints in civil engineering composites via numerical and thermographic results pre-processed by a groundbreaking mathematical approach

“…13). Both the true positive and false-positive rates are computed for different values of the threshold τ true positiverate (TPr)= true positive actual ground truth pixels (12) false positiverate ( FPr)= false positive actual ground truth pixels (13) The ROC curve is a non-decreasing curve that goes from (0, 0) to (1, 1) as the threshold τ varies.…”

“…1, not much research has been conducted so far by considering the three cornerstones, (1) IRT -(2) pre-and post-processing -(3) technical physics (i.e., applied thermal engineering). Also, only in the papers [13], [17], and [26] the applied numerical modelling was used as a mean to proceed further with experimental analyses. Readers can refer to the column on the right side in Diag.…”