Multi-Domain Joint Approach for Feature Extraction of Aerial Rotorcraft Targets

Rotary targets feature extraction, as an important method for radar target identification, often faces challenges such as severe noise interference and the submergence of rotor scattering characteristics by strong fuselage scattering features. In this paper, aiming at the difficulties in extracting feature information of aerial rotary-wing targets under low signal-to-noise ratio conditions and the poor estimation accuracy, a method of joint signal domain and image domain for rotary-wing feature extraction is proposed. Based on the principles of RGB image imaging and color separation technology, the recognition and extraction of rotary-wing features are achieved. Through simulation experiments, it is verified that under a signal-to-noise ratio of 5dB, the error in rotary-wing feature extraction accuracy is less than 7%, providing theoretical support for the precise identification of radar targets.