A classification method for ground and low altitude moving targets based on multiple features

In order to accurately identify ground and low altitude moving targets, a classification method based on multiple features is proposed to classify people, vehicles, and unmanned aerial vehicles. By analyzing the characteristics of echo signals in the time domain waveform and Doppler domain energy distribution, we extract features such as Doppler domain energy distribution, velocity, radar cross-sectional area (RCS), waveform factor, time domain variance, frequency domain variance, and frequency domain entropy. These features exhibit clear separability. Additionally, we divide the velocity interval based on the specific circumstances of the target. Based on the measured echo data, the extracted feature vector is inputted into the BP neural network for classification. The experimental results demonstrate that the target classification algorithm based on the above multi-feature and velocity interval has a strong classification effect.