Abstract: At present, the method of aggregation and splitting is used for spatial data classification and hierarchical analysis. The analysis speed is not high and the limitation is large. In order to speed up the spatial data clustering speed, improve the accuracy of data classification, this paper presents a spatial data clustering method based on support vector machine, firstly, spatial data preprocessing based on wavelet multi window feature pute filtering sieving redundant data, then filtered according to the number of divided subintervals, as defining data extraction condition query data when, according to the specific characteristics of spatial data, the spatial data describing the individual collection into a series of independent groups, determine the spatial distribution of the data set, to realize spatial data aggregation. The experimental results show that the proposed method has high classification efficiency and high accuracy.