With a vast territory

complicated natural conditions

multiplicity of crop structure

small and dispersive distribution of parcel

the precision of crop distribution maps based on remote sensing imagery can’t satisfy the need of crop yield forecasting. For this study

the paper using QuickBird high spatial resolution satellite imagery created detailed crop pattern map in a test site of Taigu

Shanxi province

where the crop pattern is very complexity in autumn. First

the QuickBird image was divided into segments by using object-oriented image segmentation technique. Second

the main land cover was classified by using spectral

spatial and contextual information based on fuzzy logic. Finally the detailed crop distribution map with high accuracy was made by combining the classification result and the field investigation. In spite of the high spatial resolution of the QuickBird BIRD image

classes such as different crops are still fairly difficult to identify. So the field investigation is very important. The map provides a more accurate spatial pattern of crops

and is useful for crop yield forecasting.