Defects Detection in Magnetic Particle Inspection Application Using Image Processing Techniques

摘要

Nondestructive testing (NDT) is widely used in many fields, particularly for criticalrnapplications such as welds of pressure vessels, ships, aircraft etc. Magnetic ParticlernInspection (MPI) is one of the best-known and commonly used methods of NDT. Itsrnaim is to detect the presence of surface braking discontinuities in the part underrninspection.rnCurrently, defects detection in MPI is still based on the subjection judgment by humanrnoperators. It is time and manpower consuming works. In addition, human interpretationrnis very subjective, inconsistent and sometime biased. So we propose a target defectrnidentification approach that is capable of automating extraction the defects in MPIrnimages. By using this approach we can develop the defect detection system in MPI fromrndepending on the human which detects the defects manually upon his experience andrnskills which varies from one to one to the automated system depending on the computerrnvision.