Abstract: Based on the UAV technology, the comparative analysis of support vector machine（SVM） and random forest（Random Forest） models was carried out with the rice yield of Nongjiang Farm of Jiansanjiang Sub-bureau as the research object. The models were evaluated by five key vegetation indices, such as R~2（coefficient of determination）, MSE（Mean Relative Error） and RMSE（Root Mean Square Error）, to analyze the performance of each model in the correlation analyses of LAI（Leaf Area Index） and above-ground biomass, as well as the correlation analyses of vegetation indices and rice yields at different fertility periods. The results show that the support vector machine model exhibits a high level of prediction accuracy, indicating that the rice yield estimation model based on UAV technology has obvious advantages in both prediction accuracy and stability, and the results are intended to help agricultural managers to make precise decisions to provide reference and reference.