Applying the good nonlinear classification ability of the least squares support vector machine (SVM) algorithm, this paper conduced the classification of land use in agricultural district from the high-spatial resolution ZY1-02C remote sensing images, which was based on the SVM method integrating information of shape and texture. It shows that the high-spatial resolution ZY1-02C data can realize land classification quickly and effectively, and the classification accuracy is increased by adding the feature information. The least squares SVM classification results were ideal, the overall accuracy was 82.53%, and the Kappa coefficient was 0.8071. It has higher accuracy than traditional method and provides a feasible method for the classification of land use based on domestic high-spatial resolution satellite.