Research On Automatic Interpretation Technology Of Winter Wheat Spatial Distribution Based On Random Forest Algorithm

Food security is the guarantee of national survival and social development,and it is indispensable at the strategic level of national security.Winter wheat is an important part of China's grain crops and an important part of grain control.The cultivation and production of winter wheat is not only related to the local agricultural development and planning,but also to a certain extent will affect the national grain output and security.Therefore,using remote sensing technology to rapidly and accurately grasp the spatial distribution situation of the winter wheat can provide important reference for the local industry planning information.This paper realizes automatic interpretation of spatial distribution information of winter wheat based on random forest algorithm,which involves three innovative points: first,the method of feature selection and multi-temporal identification;second,binary processing of the prediction results of the algorithm;third,effective identification and extraction of winter wheat planting trivial areas.The operation flow of this study is as follows: Firstly,the study area and sample area of winter wheat are selected from remote sensing images based on field investigation,and several recognition features of winter wheat are constructed.Then,the corresponding feature images are obtained by processing remote sensing images.After format transformation,the data of sample area are aggregated into training data sets.Then the test data set is obtained in the same way,that is to extract the data of spatial distribution area of winter wheat.After multi-stage image feature optimization,the random forest algorithm is used to extract the information of winter wheat.The extracted spatial data is restored and then loaded into the remote sensing image processing software to display.Finally,the automatic extraction result of remote sensing satellite image based on random forest algorithm is obtained.When the winter wheat is extracted,the collected remote sensing information is analyzed,and the extraction results are displayed on the remote sensing images with different colors or symbols,which can provide the necessary materials for subsequent accuracy verification and thematic maps.The experimental results show that the method has good extraction effect,and has certain practical application and reference value.