Abstract
The paper proposes the solution to the problem of determining the relationship between the accuracy of classification of radar images of targets belonging to ten classes (from the public portion of Moving and Stationary Target Acquisition and Recognition standard data set) and the signal-to-noise ratio (from –20 to 20 dB). Unlike the traditional approach to solving the above problem based on statistic analysis methods, the classification accuracy was estimated using an eight-layer deep convolutional neural network. It is found out that target recognition failure (a decrease in the classification accuracy down to 30 % and lower) regarding sub-meter radar imagery of motor and armoured vehicles requires interference with a background level higher than the average background level of target echoes by at least 10 dB. The study also proves that using pre-trained networks with the MobileNetV1 and Xception architectures fails to improve the classification accuracy.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Journal of «Almaz – Antey» Air and Space Defence Corporation
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
