Micro-Doppler Extraction of Cone-Shaped Target Based on STFT-RANSAC

As an important feature for micro-motion targets, micro-Doppler is widely used in parameters estimation and target classification. Aiming at the extraction of micro-Doppler, a method based on STFT (Short Time Fourier Transform, STFT) and RANSAC (Random Sample Consensus, RANSAC) is proposed. Firstly, the precession model of coneshaped target is established, and the parameterized expression of micro-Doppler is derived. Secondly, STFT is conducted on the echo to analyze the characteristic in time-frequency domain, and time-frequency maxima is used to acquire the time-frequency points set. Then, according to the parameterized expression, the prescribed fitting function is designed, and RANSAC is adopted to distinguish the valid micro-Doppler frequency points and the invalid points, and thus the micro-Doppler is estimated. Finally, the effectiveness of the proposed method is verified by experiments.