To improve the estimation accuracy of rainfall spatial distribution based on the gauge network, an interpolation method is proposed based on the matrix factorization. By combining the traditional interpolation methods, including the inverse distance weighting (IDW) method and the ordinary Kriging (OK) method, and the FunkSVD model, this method considered the temporal development of precipitation and the spatial distribution of gauges. The daily observation data of rainfall events from 2009 to 2012 between Xiaolangdi and Huayuankou were selected for the accuracy test. The results show that through combining with the FunkSVD model, the estimation errors of the IDW and OK can be reduced over 15% in terms of MAE and PERC, and the improvement is especially obvious when the surrounding stations are unevenly distributed, or the rainfall is heavy. The proposed method can greatly reduce the estimation error of traditional methods in the precipitation interpolation process and help improve the accuracy of spatial estimation.