Marine Controlled Source Electromagnetic Method (MCSEM) is an effective geophysical exploration method for detecting buried objects, mineral resources and oil and gas resources. The acquisition of marine controlled source electromagnetic signal is interfered by seawater disturbance noise, air wave and random noise, which seriously affects the qualitative analysis and quantitative inversion interpretation of electromagnetic detection data. Conventional denoising methods greatly depend on input parameters, which greatly limits the denoising effect of MCSEM data. This paper adopts Recursive Least Square (RLS) adaptive filtering method and Fast Independent Component Analysis (FastICA) method for signal separation and denoising of full-depth marine controlled source electromagnetic data. The simulation results show that the separation and denoising effects of FastICA method are better than those of RLS method, and the source signals can be better estimated from the observed mixed signals.
