In order to improve the accuracy of crop identification and expand the scope of data application, two pieces of GF-1 data with 16 m spatial resolution and 23 pieces of MODIS data with 250 m resolution were used as data sources. First, the supervised classification of GF-1 and the construction of MODIS NDVI time series were carried out. Then, the decision tree model was constructed by using the supervised classification results and time series data characteristics to identify the main crops in Shijin Irrigation Area. The overall accuracy of extraction reaches 93.13%, which is more than 10% higher than the result of supervised classification. The results show that the method can not only recognize the spatial and temporal distribution information of crops, but also give full play to the advantages of the data.