LAUSR

Effective cathode diameter (ECD) is an important index of low-light-level (LLL) image intensifier. The traditional evaluation method of this index is based on subjective observation. Professional testers observe the output image of image intensifier through microscope, and give the evaluation results according to the calibration target and their visual perception. The drawback is that the tester has to zoom the effective field of view and rotate the scale target to make evaluation, so the subjectivity is strong and the manpower is considerable. To address these problems, an objective evaluation method of ECD of LLL image intensifier is proposed, which realizes the digital measurement of the index. This method mainly composes of three parts: extracting effective anode edge (EAE), locating effective cathode center (ECC), and calculating ECD scale length. The anode output image of image intensifier is the target image. The specific implementations of this method comprise: first, extract EAE from the target image using the proposed edge extraction strategy. The square area containing the target central square (TCS) is the region of interest (ROI). Then, use the edge corner detection strategy based on neighborhood edge slope features to extract edge corners from ROI edge image. TCS corners are distinguished from other corners by their salient features, which are utilized to generate ECC. The pixel length of ECD is calculated by EAE and ECC. After that, compute the scale length of the circle diameter where the scale is located by using the proposed basic scale positioning strategy, and the conversion unit between pixels and millimeters is derived. Finally, deduce the scale length of ECD relying on the “pixel/mm” conversion unit and the pixel length of ECD. To verify the performance of this method, different types of image tubes are adopted for experiments. The results show that this method is robust to tube type, image brightness and effective anode field of view, and the calculated key parameters are in good agreement with the actual results. In addition, by comparing with the subjective evaluation method, the feasibility of this method is investigated. The results show that the objective evaluation results are consistent with the subjective judgment results, with a higher accuracy (0.001 mm). Moreover, the repeatability experimental results demonstrate that this method is more stable than the subjective evaluation method (the maximum deviation of evaluation results is 0.08 mm). In general, the presented objective evaluation method can be regarded as a reasonable alternative to the traditional ECD evaluation method for LLL image intensifier.