Interferometric synthetic aperture sonar (InSAS), a new high-resolution three-dimensional imaging device, is broadly utilized in underwater exploration. Interferogram noise removal is a necessary stage in InSAS image processing. The coupled partial differential equation (CPDE)-based method has been frequently used for image denoising. But the limitation of the traditional CPDE methods is that the edge strength function does not change over time, which leads to the convergence time becomes longer. This article puts forward a modified coupled partial differential equation (MCPDE) method for InSAS phase noise suppression that addresses the shortcoming of traditional CPDE methods. In the proposed method, a time-varying function is introduced to control the diffusion strength in edge regions. As the iterations progress, the function decreases, which guarantees the filtering intensity becomes weaken so that the detailed information will not be lost. Numerical simulations and real InSAS data experiments verified the performance of the proposed model.
🌊
