Research On Defects Detection Technology Of Pulse-excited Infrared Thermal Image

Nondestructive testing(NDT)is an essential and effective tool for industrial development,which reflects a country's industrial development level.Advanced NDT technology will promote the equipment upgrading of key components.Nondestructive testing(NDT)is a direct means to guarantee the manufacturing quality and operation safety of major equipment,such as electric power,petrochemical,transportation,aviation,aerospace,military research,etc.It has brought significant economic and social benefits.For the plate materials with various defect types,the conventional detection technology is not competent.As the thermal excitation source of pulse method contains multiple frequency components,defects of different depths can be detected at one time,so it is suitable for detecting defect components with different characteristics.Thermal excitation by light pulse is the mainstream thermal excitation method in the international application of NDT.The technology has been gradually applied in the detection of key components of the space shuttle,such as the leading edge of the wing,the reinforced composite material panel of the nose,and the turbine blade of the engine.However,in China,there are few theoretical studies on the defects of pulse excitation NDT,so it is necessary to further explore the key defect detection technology of pulse excitation infrared thermal image.Based on the physical characteristics of infrared thermal wave nondestructive detection image stimulated by pulse,this paper studies the non-uniform heating noise which affects the detection effect most.According to the signal characteristics,a background fitting method based on LM(levenberg-marquardt)algorithm is proposed,and the "pure" heating non-uniform noise signal is extracted.Combining with the mathematical model of the background image,the accurate heating non-uniform noise background image is obtained.The LM-Background Subtraction method is used to remove the uneven heating noise signal.The fitting function is constructed and the Trend Fitting method of image sequence is proposed according to the variation trend of image signal.The LM-Background Fitting method is applied to the defect detection of pulse infrared image sequence.Based on the characteristics of heat conduction and thedistribution of heat source,the Modified Edge Detection algorithm was proposed.The algorithm corrects the defect edge position and verifies it.