System of automatic recognition of aerial targets CLASSES

摘要

The utility model aims to improve noise immunity class recognition procedure objective. This result is achieved in that the system includes pre-processing unit of the radar signal detection unit, the segmentation block, selecting attributes selection and classification block prioritization support and guidance unit estimates the radius of curvature of the trajectory to maneuver portion and an indirect estimation target mass and block recognition target class is estimated curvature of the trajectory parameters and the target weight.; In the system input receives a radar signal. Provided its detection in noise, digitization, segmentation. It is also carried out measurement of target motion parameters. Unlike the prototype measured radii of curvature of the trajectory to target sites and maneuver carried indirect evaluation of its mass. Recognition objective class is carried out on the basis of two alternative methods. First, based on complex digital signal processing pre-segmented. Secondly, based on the comparison of parameters obtained at block estimates the trajectory of radius of curvature in the area of ​​maneuver and indirect evaluation target mass, with the data stored in the database. The final selection of the target class is based on the recognition of the reliability of the maximum criterion.; Checking the effectiveness of the system air target class automatic recognition performed by a natural experiment, which showed that the use of the inventive system allows to increase the probability of correct recognition of classes of targets in heavy ECM environments by 15-20% compared to the prototype. 1 n.p.f., 1 fig.