IMPROVING THE ASTRONOMICAL IMAGES PROCESSING ALGORITHM USING LUCKY IMAGING

Abstract

Lucky Imaging is one of the methods of combating atmospheric effects based on the principles of speckle interferometry. Its essence is that, since the process of atmospheric turbulence is stochastic, from the series of images obtained, some photos will be of better quality than others, that is, they will contain information at the level of the diffraction limit. Only the best images are selected, matched and averaged, resulting in an image of an astronomical object with significantly improved angular resolution [1; 3; 4; 10]. The main drawback of the classic Lucky Imaging method is that it selects only whole images of sufficient quality for averaging, the chance of obtaining which decreases with the increase of the aperture of the telescope. Also, this method does not combat the characteristic distortions caused by atmospheric turbulence, which significantly affects the quality of the obtained image, especially for astronomical objects with a large angular size, such as the Sun or the Moon. An improved algorithm for processing astronomical images has been developed on the basis of Lucky Imaging techniques. With its use, statistical methods of astronomical image processing make it possible to obtain images of astronomical objects on relatively cheap equipment, which in terms of quality and informativeness are not inferior to images obtained on large and expensive telescopes using adaptive optics or other methods of combating the phenomenon of atmospheric turbulence. Thus, even amateur astronomers get a tool that allows them to make observations on a par with the world's largest observatories. The article describes an improved algorithm for processing astronomical images, which takes into account the peculiarities of atmospheric turbulence and corrects the distortions caused by them. The algorithm itself is quite universal and suitable not only for astronomical images, but this article shows the work with images of the Sun obtained at the Astronomical Observatory of the Ivan Franko National University. The algorithm was implemented in the Python programming language using the OpenCV library (Open Source Computer Vision Library).