Integrated structural defects are one of the most common defects in the microchannel plate fabrication and physicochemical treatment stage of low-light-level (LLL) image intensifiers. Its appearance seriously affects the imaging quality. The traditional evaluation method of such defects is solely based on visual observation, and its main disadvantages are strong subjectivity and large uncertainty. To address the above problems, two objective evaluation methods of integrated structural defects of an LLL image intensifier are introduced. The first is based on local image patch extreme, and the second is based on frequency fast Fourier transform. The basic idea of the first method is to scan the effective area (EA) of the target image globally and then extract the target defect by adopting a series of local image patch extremes adjacent in the Y-axis. Experimental results reveal that this method shows excellent performance in detecting and positioning accuracy while the time cost is satisfactory. The general steps of the second method are to scan the EA globally, and then perform a fast Fourier transform on the obtained image patch; through bandpass filtering in the frequency domain and inverse Fourier transforming the convoluted image, structures or objects projecting the similar characteristics of integrated structural defects are detected. This method not only maintains good detection integrity but also ensures high detection accuracy. Through experiments on several defective image intensifiers and compared with traditional relevant technologies, we conclude that the proposed methods can realize the high-precision objective detection of integrated structural defects of LLL image intensifiers and can be used as effective alternatives to the traditional evaluation method of such defects.
Read full abstract