Research on Cultivated Land Grain Monitoring Based on Time Series Radar Data
Taking Jiexi County,Jieyang City,Guangdong Province as the research object,this paper used the time series Sentinel-1 data of the second half of 2021 to anylize the time series backscattering coefficient characteristics and inter-class differences of different covers on cultivated land such as rice,corn,pond water surface,uncultivated,woods and vegetables in the measured samples.The results showed that the optimal polarization mode of crop classification in Jiexi County was VH polarization,on which,a model for the classification of culti-vated land crops based on long short-term memory network(LSTM)was constructed.The model accuracy reached 90%.The spatial distribu-tion of rice,corn,pond water,uncultivated land,forests and vegetables in the study area was extracted based on the model,providing a new remote sensing technology for monitoring cultivated land types in cloudy areas.
cultivated land grain monitoringSentinel-1time serieslong short-term memory networkJiexi county
万方数据
维普
