WAYS OF EXTRACTION OF MAXIMUM INFORMATION FROM IMAGES OF ASTRONOMICAL OBJECTS

摘要

Generally, the information obtained as a result of a physical experiment proves to be partially corrupted due to action of various noise factors. The nature of these factors is dependent on the experimental conditions however, approaches to reducing their influence prove to be rather similar. The investigations carried out by the team of image processing of the Institute of Radio Physics and Electronics, National Academy of Sciences of Ukraine (IRE NASU) over the last ten years were oriented toward search for methods of extraction of maximum information from astronomical observations and space experiments. The investigations were conducted in several directions. For observations from the Earth's surface, the problem consisted in removing atmospheric phase distortions and synthesizing antenna configurations for telescopes and interferometers operating in various ranges of the electromagnetic wave spectrum. In the case of space astronomy, the efforts were concentrated at optimal interpretation of their results, in particular, at obtaining the planet relief and image processing with the effect of gravitational lensing. The common feature of the developed techniques is the use of the statistical approach, i.e., application of the mathematical statistics and theory of optimal statistical solutions. The present paper discloses both the general features of its approach and specific features of the obtained solutions in the case of particular problems of astronomical image processing. As a result of the performed investigation several techniques have been developed, including the method of Fourier component total phase accumulation for suppression of the atmospheric phase noise, the technique of optimum estimation of the point source intensity observed through the gravitational field of a distant galaxy, the method of accounting the altimetric information in photoclinometric determining the planet surface relief, and a number of techniques of designing antenna configurations for telescopes and interferometers. The statistical approach to processing experimental data has allowed using the information contained in a series of observations of an astronomical object for reconstruction of a single but more accurate its image. Owing to this it has become possible to unite in an optimal way the information contained in the surface images and altimetric data to obtain the surface relief with a high resolution and referencing to actual altitude magnitudes. In the case of treating gravitational mirages it has become possible to separate the observed source from the background. The obtained results demonstrate that the Bayesian statistical approach is a powerful tool in the researcher hands.