Early extraction of winter wheat planting area based on Sentinel-2 images
[Objective]This study aims to explore a method for early and rapid extraction of winter wheat planting area,rapid mapping and accuracy verification of the winter wheat spatial distribution based on high-resolution multispectral remote sensing images,providing information for high yield and good quality planting and fertilizer and water management of winter wheat in Shandong Province.[Method](1)Sentinel-2 remote sensing images were preprocessed firstly,and then the winter wheat identification sample database was constructed by combining the automatic extraction of historical planting distribution data and manual selection.The samples were divided into five categories including wheat,woodland,water body,buildings,roads and other crops.(2)The winter wheat planting area was extracted by random forest machine learning classification combined with manual interpretation of constitutional images based on the preprocessed Sentinel-2 remote sensing images,so as to realize rapid extraction of winter wheat planting area at an early time and verify the accuracy.[Result](1)The results showed that the winter wheat planting area in the study area was extracted as 544,100 hm2 with the overall distribution accuracy of 97.05％and the kappa coefficient of 0.941 0.The extraction effect of the established method was good.(2)The method proposed in this study could achieve high-precision extraction and rapid mapping of winter wheat planting area.[Conclusion]Accurately grasping the planting area and spatial distribution information of winter wheat in the early stage can provide a scientific basis for local governments and agricultural departments to arrange and guide agricultural activities.
NETL
NSTL
万方数据
