Objective evaluation of the resolution of low-light-level image intensifiers based on fast Fourier transform

Abstract

Resolution is an important index for evaluating the performance of a low-light-level image intensifier. The traditional method of image tube resolution measurement is to observe the resolution target on the image intensifier through a microscope and give a subjective judgment. The disadvantage is that the difference in the visual acuity between testers will lead to variability in measurement results. In addition, the process of focusing the resolution image is very time-consuming and eye-consuming. To solve these problems, an objective evaluation system of image tube resolution based on fast Fourier transform (FFT) that provides an efficient and feasible objective evaluation scheme for the estimation of image tube resolution is proposed. In this system, the resolution target is first focused on the cathode surface of the image tube through an optical system, and then the image is taken by a high-resolution camera. The region of interest of the collected image shows that the gray-scale sequence along the direction of the stripe change reflects the frequency of the stripe. When the gray-scale sequence is transformed to the frequency domain using FFT, the clarity and resolution of the stripe will be correlated with some quantities in the frequency domain. We extract the clarity-resolution values from the frequency domain of all stripe elements and use the generated clarity threshold combined with the linear fitting strategy to achieve the resolution value. The experimental results show that, in terms of accuracy, the test results of the system are consistent with the subjective evaluation results. For the tubes with resolutions between 61.37 and 66.49 lp / mm, the accuracy of our system is higher than that of human judgment. For repeatability, the measurement results of the system are in good agreement with the subjective evaluation results of the tubes with resolutions between 50 and 62 lp / mm. Therefore, our system can be regarded as a reasonable alternative to the subjective evaluation method, which will greatly reduce the variability caused by different testers.