Study On Image Processing Methods Of Leakage Diagnosis Of Embankment Dam Dased On The Wavelet Transform

In order to further improve the analytical accuracy of leakage diagnostic image of earth-rock dam, this thesis combined the processing methods of image denoising based on wavelet transform,keeping the image information the most,with the NDT method of leakage diagnosis of earth-rock dam based on resistivity tomography and wave velocity tomography to analyze leakage diagnostic image of earth-rock dam jointly.Meanwhile, software that process the leakage diagnostic image of earth-rock dam,was developed (Hereinafter referred to as the software).The software could be qualitative and quantitative analysis leakage diagnostic image of earth-rock dam,which provided theoretical basis and application reference for processing methods of leakage diagnostic image of earth-rock dam.The main research contents and results were as follows:1.According to characteristics of multi-phase mixture of soil-stone in the water conservancy engineering discipline, leakage factor of earth-rock was established based on resistivity tomography.According to the equivalent elastic characteristics of multi-phase mixture of soil-stone and theoretical formulas for calculating velocity, leakage factor of earth-rock was established based on wave velocity tomography.2.According to image analysis theory in the information technology subject, parameters described accuracy of leakage area characteristic were proposed.Such as regional circumference L;regional area A; length-width ratio of circumscribed rectangle E and compact degrees F.3.According to analyzing generation mechanism of noise,white Gaussian noise model were utilize.4.Wavelet transform were applied to denoise digital image with white Gaussian noise by using the Donoho and Johnstone definite threshold and soft threshold function,and Kirsch operator were applied to withdraw edge characteristic information.5.According to international universally available evaluation criteria,subjective assessment method and objective evaluation criteria were quoted. Universally available subjective assessment criteria(Impedence with scale level 5 scoring method and Quality with scale level 5 evaluation method) and objective evaluation criteria of professional image assessment criteria(Standard deviation of image; peak signal noise ratio,information Entropy,average grads,edge strength)were applied to assess quality of the image which had been denoised.6.The software utilized the numerical model experiment of leakage diagnosis image of earth-rock dam to verify and assess quality of the image which had been denoised by methods of mean denosing;clustering analysis and threshold filtering,and digital image characteristic information of numerical model of leakage diagnosis image of earth-rock dam which had been abstracted by using Kirsch operator.According to analyzing by the software,the conclusion that threshold filtering had a good effect on denoising was drawn.Kirsch operator was applied to evaluate and analyze the leakage disease area of numerical model of earth-rock dam with high precision.By comparing,the diagnostic results of the crack disease region and shape and area of round hole disease region on Noisy theoretical model and Theoretical Model of earth-rock dam were in good agreement,and the error range was within 0.7%.7.According to comparative analysis of the leakage diagnostic results of physical model experiment and numerical model experiment of earth-rock dam,the conclusion, which combining the processing method of image denoising based on wavelet transform with the NDT method of leakage diagnosis of earth-rock dam based on resistivity tomography and wave velocity tomography to detect the crack disease region of earth-rock dam jointly was feasible,was drawn. The shape of crack disease region was described in much deviation.But detection of the round-hole type of disease region was feasible with high precision.Therefore, image processing methods of leakage diagnostic image of earth-rock dam based on wavelet transform would improve analytical precision of leakage diagnostic image of earth-rock dam.