Bathymetric points are the main data source for obtaining high-quality underwater Digital Elevation Models (DEMs). The low density of bathymetric points in coastal zones can lead to low underwater DEM accuracy. To address this problem, we propose a multiparameter collaborative optimization algorithm for point densification that considers spatial anisotropy. First, the particle swarm optimization was employed to collaboratively optimize the four parameters affecting the computational accuracy to achieve the overall tuning and determine the optimal solutions of the four parameters. Subsequently, the four determined parameters were applied to the inverse distance weighting method to obtain in the global sense the optimal value of the complementary point depth. Finally, the proposed densification method was validated using four common interpolation methods in geosciences and bathymetric data from the coastal zone of the Yellow River Delta from 1992, 2007, and 2015. The experimental results showed that this method significantly improves the interpolation accuracy of the processed data, reducing both the absolute and relative errors by 12%. This method overcomes the problem of large interpolation errors caused by the density of bathymetric points and allows for a more accurate underwater topography reconstruction.