Study of Seismic Denoising via Decomposition Algorithm Based on Windowed Frequency Modulated Atom

This paper studies the seismic denoising algorithm based on windowed frequency modulated atom decomposition principle. Adopting the decomposition algorithm based on the simplest windowed frequency modulated atom - Gaussian Chirplet to decompose noisy seismic signal, thereby extracts the valid signal components to achieve the purpose of denoising. Although it does not provide the premise that statistical characteristic of background noise of seismic signal, this method could attenuate the noise effectively. Through some simulations on seismic records, it demonstrates that the algorithm recovers the valid data contaminated by the noise very well.