Estimating the probability of flaw detection by modeling processes of image generation and interpretation

Abstract

A method for determining the probability of flaw detection by modeling optical images generated by X-ray nondestructive testing methods is described. Processes of formation of intrinsic noises in transducers have been studied. Normal distribution laws have been substantiated for the optical density in radiographic evaluation and the image brightness in fluoroscopic testing. Based on the image recognition theory, application of convolution in X-ray image interpretation has been justified and the accompanying change in the signal-to-noise ratio has been demonstrated. Dependence of flaw-detection probability on the shape and dimensions of test articles has been studied.