Dense Dark Vegetation algorithm (DDV) is effectively used in retrieval of aerosol optical depth (AOD) from satellite images over land. However, the algorithm is only suitable for low albedo areas such as water and dense vegetation areas. The purpose of this paper is mainly to retrieve aerosol optical depth over none DDV pixels by using the kriging interpolation method, we first retrieve the aerosol optical depth from EOS-MODIS data (a resolution of 1km∗1km) by the DDV algorithm, but there are many missing values distributed on the image. Then we interpolate the AOD image and get values of none DDV pixels with the kriging interpolation method. Our results show that the kriging interpolation results are consistent better with the surface observations than traditional interpolation methods and produce a better visual effect than the original image.