搭载多幅图像平均分辨率自动检测装置的直视型微光夜视仪测试研究

Test and research of direct vision LLL night vision instrument equipped with automatic detection device of multi-image average resolution

当前夜视仪及众多光学仪器的分辨率检测正朝着更智能的方向发展。传统的人工检测由于人眼的差异等因素会导致检测结果出现误差,还存在检测效率低的问题。为此,研究提出了一种搭载多幅图像平均分辨率自动检测装置的直视型微光夜视仪。研究通过Hough变换设计了分辨率图案的定位算法,将对比度作为分辨率阈值对图像分辨率进行自动判读,并采用多幅图像平均法与形态学运算来提高定位算法的精度。结果表明,当平均图像数量逐渐增加时,定位精度迅速提高,特别是其数量增至10幅时,定位精度达到1.9个像素。研究的搭载多幅图像平均分辨率自动检测装置可以实现左、右目分辨率检测,并且该检测方法更稳定、更准确。

At present,the resolution detection of night vision and many optical instruments is developing towards a more intelligent direction.Due to the difference of human eyes and other factors,the traditional manual detection will lead to the error of the detection results,and the problem of low detection efficiency.To this end,a direct light level night vision instrument equipped with automatic detection device of multi-image average resolution is proposed.In this paper,the Hough transform is used to design the positioning algorithm of resolution patterns.The contrast is used as the resolution threshold to automatically interpret the resolution.The average method of multiple images and morphological operation are used to improve the accuracy of the positioning algorithm.The results show that the positioning accuracy improves with the increase of the average number of images.When the number of images reaches 10,the positioning accuracy reaches 1.9 pixels.The proposed automatic detection device equipped with multiple image average resolution can better detect the left and right eye resolution of the system to be inspected,and is more stable and accurate than the current detection method.