As one of the major crops in the world,the spatial distribution information of winter wheat plays an important role in monitoring winter wheat growth,assisting economic decision-making and addressing regional food security under climate change.This paper proposed a new anti-noise identification method for winter wheat identification based on the 250m MODIS-NDVI time-series dataset during the period from September 30,2014to June 26,2015.With the method,the spatial distribution of winter wheat in Henan province was extracted based on the analysis of winter wheat phenology.Results indicated that the total identification accuracy of winter wheat was 93.0%,94.0%and 86.0%for the whole study area,fragmentary land area and regular land area,respectively.Compared with the traditional identification method for winter wheat based on satellite time-series data,the identification accuracies with the proposed method in different filtering scenarios were not only high but also similar to each other.It strongly proved that the new method had a good performance in noise immunity and stability and can be applied to the rapid extraction of winter wheat in a large scale based on satellite time-series dataset.This new method provided a new technical support for the operational extraction of winter wheat.