Robust multistage Doppler frequency extraction for laser velocity measurement under low SNR conditions

This paper proposes a novel hybrid Doppler frequency estimation method for Laser Doppler Velocimetry (LDV), designed to improve measurement accuracy in harsh industrial environments, such as steel production plants. The proposed system integrates autocorrelation filtering, wavelet denoising, Zoom-FFT spectral refinement, bandpass filtering, and zerocrossing detection. According to simulation results, the proposed method achieves a mean absolute velocity error of less than 0.005 m/s under low signal-to-noise ratio (SNR) conditions (target velocity: 30 m/s), and outperforms conventional frequency estimation approaches including Fast Fourier Transform (FFT), Continuous Fourier Transform (CFT), and energy centroid correction. The effectiveness and robustness of the proposed approach are validated through synthetic LDV signal experiments, demonstrating its suitability for high-speed bar and wire measurement applications.