Abstract
The article describes methods for extracting valuable data from echo signals received from natural and artificial aerial objects recognized by mobile radar stations. Options for filtering the signal reflected from air objects, presented in a complex form, are considered. A method for identifying signal features based on radial basis functions is developed. A hybrid model of the classifier for the recognition of unmanned aerial vehicles and ornithological objects is proposed. It includes subsystems for filtering, feature detection based on clustering and approximation functions, as well as the classifier itself. Methods of recognition based on neural networks are considered. To train the classifier was used a library based on the principle of gradient boosting of decision trees. An experimental sample was used to train and test the classifier.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have