Resolution is one of the important indicators that must be evaluated for low-light-level (LLL) image intensifiers. The traditional techniques for evaluating this parameter include visual observation and objective testing. The visual evaluation method has the disadvantages of strong uncertainty and low accuracy, whereas the existing objective testing techniques mainly suffer from poor robustness. To address the above problems, an automatic evaluation method for this parameter based on frequency spectrum analysis of a two-dimensional (2D) matrix of stripe pattern is proposed. First, a 2D matrix of stripe image with stripe pattern position displacement as X coordinate and stripe clarity change as Y coordinate is established and converted into the frequency domain. Then, the critical value of stripe clarity is determined by combining the brightness variation curve of the spectrum feature points of the precisely located stripe images in the matrix with the extremely discernible stripe image. Finally, this value is applied to the spectrum feature point brightness change curve of the matrix established from the stripe unit image to obtain the image tube resolution. To substantiate the performance of this method, four state-of-the-art techniques are compared, and several LLL image intensifiers are put into the experiment. The results indicate that the proposed method has the most consistent results with the visual evaluation method and manifests the shortest testing time. The applicability verification experiment results demonstrate that the repeatability of this method is satisfactory, and the probability of fluctuation not exceeding one stripe unit is not greater than 20%.