Application of multi-threshold adaptive filtering in noise processing of fishfinder

In joint geophysical exploration mission, strong and banded interference noise in the fish finder data for research are is common, which could seriously affect the accuracy of the identification of seafloor plumes and the like. To improve the signal-to-noise ratio of the profile and the fidelity of the water structural information, we proposed a multi-threshold adaptive filtering method that integrates the spatial features of target strength, orientation, and gradient change, constructed a triple threshold determination system, and introduced a sliding window adaptive median filtering mechanism, with which the banded noise can be suppressed, the boundary clarity of the seafloor cold-spring plume and the continuity of the internal structure be enhanced, The practicality of the method was verified by modelling, and good results were achieved against the measured cold spring data. This study provided a reliable pre-processing basis for the subsequent identification of water body structure and quantitative analysis.