Research Of Algorithm For Achieving High-Frequency Images From Low-Frequency Scanning Images And Tomography Processing To Focal Plate Thermal Wave Images In Infrared Thermal Wave Nondestructive Evaluation

In this paper the principle, the history and domestic and overseas development of infrared thermal wave nondestructive evaluations (NDE) are introduced. And the choice of fitting data, fitting time and the method to fit as well as the common image processing method of gray scale transformations, smoothing, sharpening, edge extraction are also introduced. Then structure and algorithm of the processing program of the optical mechanical scanning image are introduced. At last the results of processing to the simulative data and the real data are given.The paper proposes two kinds of fitting processing method, the horizontal fitting processing and the vertical fitting processing, which are aimed at the optical mechanical scanning image in different defective conditions. The optical mechanical scanning image processing program used to actualize the algorithm brought forward by this paper is coded with the MATLAB program, its functions consist of the reading of optical mechanical scanning thermal wave image data, the fitting of thermal wave image data, the determining and quantifying of changing rules of pixels, the optimizing of displaying the effect of the image, reconstructing the similar focal-plane thermal wave image from optical mechanical scanning thermal wave image and getting high-frequency digital image at last. The result of the optical mechanical scanning image processing program show that achieving high-frequency images from low-frequency scanning images is feasible in algorithm, and the processing effect is comparatively good.The paper also proposes the algorithm of tomography processing to images data captured by focal plate array (FPA) infrared thermal images camera, proves that the lnTs-lnt curve will be deflected when thermal wave transmits from one medium to another, and the deflecting time is just the time when thermal wave reaches the edge of the two media by simplifying the thermal transmit formulae. So once given all the deflecting time of the pixels on the surface of the sample it can be determined that whether the area where thermal wave reaches is defective area or non-defective area, so as to obtain the purpose of tomography. The real processing results to fiberglass sample proves that the transverse section images in any time can be obtained clearly by using the algorithm to process the FPA thermal wave images proposed by this paper, and the position, size and deep information of the defective area are also really showed.