Methods for logical processing of images of radar objects marks based on semantic features

Abstract

This paper considers a method for logical processing of radar images based on semantic features. Algorithms and programs for automatic detection and recognition of radar aerial objects in surveillance radars with processing of real signals recordings are also given. The relevance of this work is the development of automatic information processing algorithm to ensure effective detection and recognition of radar object signals based on semantic features. The advantage of this method is the possibility to bring the image processing procedure of radar object images closer to the expert’s logic. It implies involvement in analysis of various distinguishing features between reflections from aerial objects. The problem of detecting and recognizing images of radar objects is transformed into the problem of feature classification. Therefore, the essence of logical image processing method based on semantic features is a making decision about detection and recognition of radar objects based on comparative analysis of features, which are defined on the set of semantic, fluctuation, geometric and energy components of radar image. An algorithm for automatic decision-making on detection and recognition of aircraft radar signals, including point-moving and stationary aircraft such as airplanes, helicopters, UAVs, is given. This approach of forming semantic features of radar signals, as well as the automation of information processing operations, increases the effectiveness of detecting weak signals due to accumulation of signal (energy) and logical information. At the same time, logical information is accumulated from the analysis of dynamic map of radar signal intensities with tracking of changes occurring in it during several radar soundings.