Image Analysis And Processing Of Composite Plate Defect Ultrasonic Tomography

The comprehensive performance of carbon fiber composites is superior.So it is widely used in various fields.Because the defects of the material itself or defects caused by irrational use will bring about potential safety hazard when we use it,even cause huge losses.Therefore,the detection of material defects is particularly important.Ultrasonic non-destructive testing technology is an advanced non-destructive testing technology.Computer tomography technology is deeply applied to the medical field.Digital image processing technology is promoting the development of modern digital signal processing.The combination of the three,is rather significant in the research of the defect detection of composite materials.By reviewing the domestic and foreign literatures,we understand the propagation characteristics of ultrasonic waves in homogeneous and anisotropic materials,and give the dispersion equations and dispersion curves of ultrasonic Lamb in these two different materials,which provides a theoretical basis for the extraction of ultrasonic Lamb wave excitation signals and get the time of fly in the finite element simulation and experiment.The theoretical basis of computer tomography is introduced in detail,and the mathematical formulas of parallel scanning filter back projection algorithm and fan scan filter back projection algorithm based on central slice theorem are deduced.Finally,the details of step of these two algorithms are given respectively.By contrast,the fan-shaped scan filter back-projection algorithm is selected finally for the image reconstruction algorithm used in this paper simulation and experiment.Selecting the carbon fiber composite multi-purpose plate T300/5028 as the object of this paper,the ABAQUS finite element software was used to establish the thin plate model with the layered defects in different positions.The five-cycle Henning window sinusoidal signal is used to simulate the propagation of ultrasonic lamb waves in the thin plate.We should extract the data of TOF.Through the MATLAB program to write fan scan filter back projection algorithm program,the extracted travel time data into the algorithm to reconstruct the thin film model,the defect information is clearly visible.Sharpening filtering,global threshold,convex hull connection,edge detection,connectivity selection,and image cropping are introduced.The traditional convex hull connection algorithm—Graham scanning method—is improved for the inherent characteristics of composite board defect images.Maintain higher accuracy while having faster execution efficiency.Based on the basic theory of image processing,an image processing method for defect detection of composite materials is proposed.The center defect image and skewed defect image of the composite plate were processed,and the feasibility and practicability of the algorithm were demonstrated from the experimental results.