A fast and automatic detection method for resolution of direct-view low-light night vision device
Because traditional methods rely on manual interpretation for resolution detection, differences in human eyes, fatigue, and psychological factors may lead to differences in detection results. At the same time, the manual detection method also has the problem of low detection efficiency. In order to enhance the objectivity of night vision device resolution detection and improve the efficiency of resolution detection, it is necessary to automatically detect the resolution of night vision devices. This paper proposes an automatic detection method for night vision device resolution.
1 Detection principle
The basic principle of the automatic detection of the resolution of the low-light night vision device is: firstly locate the resolution pattern, then judge whether the resolution pattern is distinguishable according to the objective criterion of the resolution, and determine the resolution of the instrument according to the result of the judgment. The resolution automatic detection is carried out according to the following steps: firstly, the image of the specially-made resolution target plate is collected with a CMOS camera, and the image is preprocessed. Then, the Hough transform algorithm is used to detect the positioning lines in the image, and the position of the center of the resolution pattern, the size of the pattern, and the inclination of the pattern in the entire image can be obtained according to the positions of the four positioning lines. Extract a resolution pattern composed of equally spaced and oriented stripes, for each pattern
Contrast calculation is performed, and whether the region is distinguishable is judged according to the magnitude of the contrast. Contrast is calculated starting from the pattern with the widest line and going down until the pattern that cannot be distinguished is calculated. For the left and right resolution patterns of the same group, the contrast ratio of the pattern with the lower contrast is used as the contrast ratio of the group of patterns. The number corresponding to the previous group of indistinguishable patterns is taken as the resolution of the inspected night vision device. If the first set of resolution patterns is unqualified, it is determined that the resolution of the night vision device is lower than that of the pair
If the last set of resolution patterns is still distinguishable, it is judged that the resolution is higher than the corresponding value of this set. Suppose the resolution target has a max set of resolution patterns.
Determination of resolution threshold
When performing resolution detection, the insufficient resolution of the system under test will blur the resolution pattern and reduce the contrast of the resolution pattern area. When the contrast of the resolution pattern area is lower than a certain value, the human eyes can no longer distinguish it, and it can be considered that the resolution pattern cannot be recognized. The contrast of an image is calculated using the following formula:
Among them, Imax is the maximum brightness of the area, and I min is the minimum brightness of the area. When reflected in the digital image, it can be replaced by the gray scale of the pixel. When studying the resolution of a diffraction-limited system, the Rayleigh criterion holds that when the center-to-center distance between two diffraction spots is just the radius of the first dark ring, the human eye can just distinguish these two points, and at this time calculate the area’s area according to the contrast formula Contrast is .
