Cropping index is a very important indicator

which reflects the situation and degree of arable land to be used by a certain planting system at a certain period. Crop growth dynamic can be monitored by the time series of NDVI data. The differences of crop growth show in the curve of time series of NDVI clearly. The curve of time series of NDVI describes the crop process of seeding

jointing

tasseling

and harvesting and so on. There exist some peaks and valleys on the curve of time series of NDVI. These peaks correspond the period of crop tasseling

and valleys the period of crop harvesting. The link that connects the cropping index with a time series of NDVI is the seasonal rhythm of agricultural crops in a year. Time series of NDVI contain the rhythm of vegetation growth and wilt. But due to cloud contamination

the curve of time series of NDVI has a lot of noise. This paper tried to remove the cloud contamination from the curve of time series of NDVI with the assistance of HANTS software. The reconstructed time series of NDVI can accurately reflect the biophysical processes of planting

seedling

elongating

heading

harvesting of agricultural crops. So according to the period of time series of NDVI

the dynamic information of crop under cultivated land can be extracted and cropping index of arable land can be further calculated. The paper presentsu remote sensing method for extracting the cropping index and then extracted the cropping index from 4 years of VEGETATION decadal composite time series of NDVI over the period of 1999 to 2002. The validation results show a high accuracy compared to the 4 test sites ground data and other available information.