On the correlation analysis of the reprocessed digital images of some space maps
Eimuntas Paršeliūnas, Dominykas Šlikas · Scientific Reports · 2026
The correlation analysis executed on some digital images of the gas spacemap of the central zone of the Milky Way galaxy is presented. The aim of this article is to develop and present a formalized correlation analysis methodology based on the theory of random functions, designed to study the spatial structure of astronomical space maps and digital images of gas nebulae and to assess their statistical non-homogeneity. This paper firstly describes the pixel vector transformation algorithm, then presents the results of the correlation analysis and provides their astrophysical interpretation. It was suggested to calculate the estimations of the auto-covariance functions of a single digital image as well as estimations of cross-covariance functions of two images on the base of random functions, which were constructed from the vectors of the digital images pixels. A method to detect the estimations of the pixels vectors is based on allocating the pixels arrays of the images into separate columns. The RGB colour model of the colours spectrum was used to code the pixels of the digital images. It is assumed, that in the case of changes of the digital image scale the colours frequencies assigned to the particular pixel remains unchanged i.e. the scale changes do not have any influence on the detection of the covariance functions. It is analysed the influence of the components of the colours spectrum and the colour’s tensor on the estimations of the auto-covariance and cross-covariance functions. Applied correlation analysis shows that the structure of the gas nebula of the central zone of the Milky Way galaxy is non-homogeneous. The scientific novelty of the article is manifested through a formalized statistical method that allows revealing the heterogeneity of the spatial structure of astronomical images using RGB color encoding, correlation analysis and the theory of random functions.