Özet:

Flaw detection of critical parts used in many sectors, especially in the automotive, aerospace and defense industry sectors, is made by Nondestructive Testing (NDT) techniques. Maximum reliability and quality of critical parts are achieved quickly and cost-effectively with NDT methods. Magnetic Particle Inspection (MPI) is widely used to detect surface and subsurface cracks in ferromagnetic materials. It also has important use in industrial applications where all critical steel parts are inspected during the production phase and at some point in service. In this study, a systematic approach is proposed with selected image processing algorithms to assist the operator in identifying defects in images taken by MPI experiments. Images taken with MPI experiments were transferred to the software and analyzed with the suggested methods. It was observed that the obtained results were compatible with the artifacts in the reference test block. Experimental studies have shown that the proposed method can provide important contributions to the operator in the interpretation of cracks. The M2GLD algorithm was able to detect all 6 defects in the reference block, while the K Mean Clustering algorithm was able to detect 5 of them.

Anahtar Kelimeler: