Helicopter Target Recognition Based on the Frequency Domain Adaptive Convolution Kernel Filtering

The characteristics of the distribution of the helicopter radar echo on the time-frequency two dimensional detection plane, a helicopter recognition method based on adaptive convolution kernel filtering in frequency domain is proposed. Comparison with most target radar target echoes, helicopter rotor echo presents special expansion characteristics on radar detection plane. Especially on low pulse repetition frequency radar, the echo of the rotor further presents of the frequency under-sampling characteristics. This feature has a greater degree of division in frequency dimension with other target echoes. Two dimensional convolution filtering of frequency domain adaptive convolution kernel and measured echo data is carried out. The filtering results can be used for echo identification of helicopter main rotor. The processing of multi frame radar echo data and statistics of helicopter target recognition results show that the method has high recognition probability for helicopter main rotor echo.