Coupling the chain classification with greedy breadth-first-search (GBFS) algorithm, a new approach has been proposed for automated sizing of defects, utilizing radial basis function neural network (RBFNN) and support vector machine (SVM). It has been established that the proposed approach can successfully size subsurface as well as surface defects.
Chain classification has incorporated dependency among the class variables (length, width, depth and height) and optimal sequence for sizing has been determined, for the first time.
Chain classification with SVM appears very good for sizing of subsurface defects, especially, the depth and the height. This algorithm has ensured a mean accuracy of 94% as compared to 91.5% achieved by RBFNN.
Chain classification has been able to significantly incorporate the dependency existing among the class variables and has successfully resulted in sizing of defects located even 3 mm below the surface.
Chain classification has been able to successfully classify, depth and height of the surface as well as near-surface defects, confirming its robustness.