INVESTIGATION OF ALGORITHMS OF DETECTING OF THE CHARACTERISTICS OF REMOTE OBJECTS IN OPTOELECTRONIC SYSTEMS BY THE METHOD OF WAVELET TRANSFORMATION

Abstract

The issues of the accuracy of measuring the coordinates and size of objects observed by an optical system by the methods of single and double continuous wavelet transform in their image are considered. It is shown that the use of the second continuous wavelet transform to the curves of the coefficients of the first transform leads to an increase in the extrema of the scalegram and the smoothness of the curves of the coefficients, providing more than two times higher coordinate sensitivity of determining the position and orientation of objects. The use of different types of wavelets in each continuous wavelet transform of signals gives many options for the curves of the coefficients of the continuous wavelet transform and can be used for additional filtering of noise, taking into account the nature of objects. The parallel use of mathematical models and real objects in a neural network for determining the coordinates of signals and their characteristics is proposed, which leads to an increase in accuracy for each type of object, the possibility of constructing intelligent control devices for outer space. Using the example of an experimental installation of two synchronously movable optoelectronic systems, the accuracy of combining images of objects in multiplexing systems from different digital sources is demonstrated.