In order to improve the accuracy and spatial coverage of AOD datasets, we proposed a method to obtain a consistent dataset with higher spatial coverage and better accuracy from Deep Blue (DB) AOD and Dark Target (DT) AOD products. The fusion algorithm consists of three parts: the first part is to remove the system errors, the second part is to calculate the uncertainty and fusion of datasets using the maximum likelihood estimate method, and the third part is to mask outliers. The MBE, MAE, RMB and RMSE of DB AOD in 2015 are 0.04, 0.13, 1.10 and 0.20 respectively, the MBE, MAE, RMB and RMSE of DT AOD in 2015 are 0.07, 0.12, 1.18 and 0.17 respectively, the MBE, MAE, RMB and RMSE of combined AOD provided by MODIS in 2015 are 0.05, 0.11, 1.12 and 0.16 respectively, and the MBE, MAE, RMB and RMSE of fusion data after mask with a threshold of 0.20 in 2015 are 0.03, 0.10, 1.08 and 0.15 respectively. The accuracy of fusion data after mask is obviously superior to the original data and the combined data provided by MODIS. In addition, the spatial coverage of the data has also been significantly improved.