Remote sensing recognition winter wheat with change vector analysis (CVA) based on cultivated parcel data

The method of CVA is widely used in the studies of the change detection between years, but there were few researches to study the remote sensing recognition crop change between a short time. This paper takes tongzhou, Beijing, as the study area. Winter wheat planting area has been acquired with the method of CVA based on multi-temporal HJ-1-A satellite´ CCD data which have a short cycle time. Further more the cultivated parcel data was used as adjust units to correct the measurement result, it has solved the registration error of multi-temporal image at a certain extent and improved the accuracy of winter wheat area measurement. There are some merits about this method. Firstly, the overall pixel accuracy of winter wheat is 91.4% and kappa is 0.801, which is higher than that acquired through post-classification comparison (which are 88.1% and 0.736).Secondly, the method of CVA is more sensitive to the spectrum change. For the growth change of vegetation in different seasons can be detected and the overall accuracy is improved, this approach can be used on the other crop area survey but not only winter wheat. Finally, the cultivated parcel data was used as adjust units to correct the measurement result, it has solved the registration error of multi-temporal image at a certain extent and improved the accuracy of winter wheat area measurement.