Численное исследование центроидирования изображения оптического излучения в плоскости регистрации

This paper presents the results of studies performed on the basis of a numerical model of the Shack-Hartmann wavefront sensor and estimating optical radiation in the focal plane using the weighted center of gravity algorithm depending on the intensity of phase distortions. The subject of the study is the algorithm of weighted centers of gravity of focal spots recorded by the photosensitive matrix of a video camera from the standpoint of the accuracy and speed of optoelectronic systems designed to measure phase fluctuations of optical radiation under atmospheric turbulence. Methods for optimizing the centroiding of the radiation image in the recording plane are considered. The results of experiments performed on the basis of a numerical model of the Shack-Hartmann wavefront sensor are presented. It is shown that estimating the product of the recorded intensity distribution and the two-dimensional Gaussian distribution as a weighting function within an analysis window of a size corresponding to the variance of the normal Gaussian distribution ensures the speed and accuracy of measuring local tilts of the optical radiation wavefront. An architecture of artificial neural networks for centroidization of radiation images in the recording plane is presented.