A method and device for object type identification

Abstract

The embodiment of the present invention discloses a method and device for target type identification. The method in the embodiment of the present invention includes: collecting the acoustic signal of the target through a planar sound transmission array, and the target includes a low-flying unmanned aerial vehicle, a light aircraft, and a power triangle. Wing, cruise missile, powered paraglider, carry out direction estimation to described acoustic signal, and carry out airspace filtering process to the acoustic signal after direction estimation, after airspace filtering process, extract the feature of described acoustic signal, The eigenvector corresponding to the feature of the acoustic signal and the feature of the target, determining the type confidence of the target according to the feature of the acoustic signal, the eigenvector corresponding to the feature of the acoustic signal, and the feature of the target , identifying the type confidence by fusion identification technology to determine the type of the target, and the embodiment of the present invention improves the accuracy of target type identification.

Description

Technical field [0001] The present invention relates to the field of detection technology, in particular to a method and device for identifying target types. Background technique [0002] Our country has a vast territory and important defending targets are spread all over the country. It is prone to attacks from armed helicopters, cruise missiles, drones and other high-tech aircraft. It is urgent to establish a low-altitude target detection and early warning system to protect important areas. The low altitude and small radar cross section make it difficult for general radar to detect, but the sound radiated by these targets during flight is difficult to eliminate. Using these sounds to locate and identify low-altitude flying targets is the two keys to passive acoustic detection that modern scholars are generally concerned about technology. [0003] Using different target types to radiate different acoustic signals, by extracting the voiceprint features of different target acoustic...

AI Technical Summary

Problems solved by technology

[0004] However, template matching technology can only be used under specific conditions. Neural network technology requires sufficient sample data, sufficient training for the network, and the convergence of network weight coefficients to the global optimal solution. Otherwise, the neural network is very difficult to use in practice. It is difficult to achieve the expected effect, so it is difficult to achieve the expected effect in actual use of the classification recognition technology in the sampling technology, which affects the accuracy of the recognition result

Images