The dissertation is devoted to solving the scientific problem of increasing per-formance of optoelectronic surveillance systems (OESS) with image fusion by creat-ing scientific bases and applied methods of analysis, synthesis and adjustment of in-formation channels on a uniform methodological basis. The methodological bases for improving the OESS power consumption with image fusion are scientifically substan-tiated. They include the methods of constructive harmonization of the OESS’s main units characteristics, the mechanism of the adaptive selection of the best method of spectral images merging, the means of experimental determination of the OESS main characteristics and allow increase performance of the system.
By information fusion from different channels, one can dramatically improve the quality of background status (BS) information. There is still no unified OESS design technology to analyze and synthesize the entire information system, to formulate technical requirements for individual blocks of spectral channels and algorithms for image fusion, taking into account the context of the observer's current task and type of BS. A number of problems remain to be solved with regard to improve the OESS performance by forming the image in operator’s field of view, which content pro-vides best solution to the current surveillance task.
An «object - OESS with image fusion – operator» mathematical model has been created. It allows analyze and design the OESS within a linear model. The formula for modulation transfer function of the “lens - detector” system of the aerospace OESS was obtained at arbitrary angles of view. It can be used to calculate the values of the lens pupil diameter, focal length of the lens and threshold irradiance of detector, which is required to achieve the given image contrast. Unified algorithm for estima-tion of spatial and energy properties of OESS spectral channels based on the targeting task performance (TTP) metric is offered. It allows to determine the main information channel in the OESS as the base for further fusion, as well as to coordinate the pa-rameters of the input blocks of two-channel OESS.
The method of performance estimation of OESS with information fusion on the basis of a posteriori TTP metric is developed. The metric is determined numerically for the method of merging images on the resulting spatial spectrum of images. This method allows to investigate the features of the image fusion process and its visual perception by the operator. It allows you to calculate the probability of detection, recognition and identification of an object, which is observed by the OESS with im-age fusion.
A new method of reconciling the basic parameters of monoblock collimator and telescopic sights for small arms is proposed. Two variants of monoblock collimator sights with improved aiming accuracy are offered. The sight of two optical parts with different refractive index, glued on the plane, provides a linear aperture of the original pupil of the sight 0.276, with parallax 3 mrad. A monoblock collimator sight of two optical parts with different refractive index, glued on a spherical surface, at the same linear aperture of the original pupil and mass provides parallax of 1.25 mrad. A scheme of a decentralized telescopic monoblock is proposed, in which at a mismatch of 3 mrad the magnification can reach 1.6 times compared to 1.4 times for an ax-isymmetric scheme.
A new method for measuring OESS spectral channels basic energy characteristics has been developed. This method, compared to its analogues, increases the dynamic range up to 2.5 times with a high uniformity of the brightness field.
For the first time, the efficiency of the image fusion strategy, which allows to increase the maximum detection range of objects in dual-channel OESS by up to 12%, was developed and experimentally validated. New methods of increasing the speed of information processing in multichannel OESS with image fusion are proposed. The first method uses the information map based on the low spatial resolution channel and the second method uses a digital coherent optical processor. The amount of computation in the first case can be reduced up to 2 times.
Keywords: optoelectronic surveillance system, image fusion, performance eval-uation, radiometric resolution, measurement of energy characteristics, spatial resolu-tion enhancement, small-scale sights, visual perception.
