Area Extraction Of Winter Wheat In Four Cities Of Huang-Huai-Hai Based On Multi-temporal Remote Sensing Data And SVM

Remote sensing technology is an indispensable means to study modern agricultural monitoring and estimation.Since the color,shape,texture and other characteristics of different kinds of crops on remote sensing images are different,these areas can be used to extract and monitor the area and growth of crops.Compared with traditional monitoring methods,modern remote sensing technology has the advantages of real-time,objectivity and wide-area,which are not available in traditional methods.Winter wheat is the second largest food crop in China,and the precise extraction of wheat area plays an important role in food production and disaster assessment.Nowadays,the rapid development of pattern recognition provides powerful technical support for the intelligent processing of remote sensing data.In this paper,Hebi City,Zhangzhou City,Shijiazhuang City and Tai'an City,the main production areas of winter wheat in Huanghuaihai,were selected as research areas.The cross-validation SVM method was used to extract the winter wheat planting area from the four-city multi-temporal Landsat satellite image,and analyzed by principal component analysis.And the multiple linear regression analysis analyzed the driving factors affecting the winter wheat planting area in the four cities.The main work contents and results are as follows:(1)Firstly,the Landsat image to be classified is preprocessed,the NDVI vegetation index map is made by ENVI,and the non-vegetation part is removed by Arcgis,and then the image mask was made by the administrative area vector and the single-band image of the green band.Get the green band vegetation map of the four cities.The image was input into MATLAB software.The SVM algorithm was used to classify the winter wheat area in different years of Hebi City,Zhangzhou City,Shijiazhuang City and Tai'an City.The SVM parameter C and parameter V were selected to obtain the optimal combination through cross-validation.The cross-validation model selects K-CV,and finally obtains the classification map and interprets the winter wheat planting area.(2)Through principal component analysis and multiple linear regression analysis,it is found that the driving factors affecting the winter wheat planting area of the four cities of Huanghuaihai are mainly the number of rural employees,urbanization rate and wheat production price index.Among them,Shijiazhuang City is most affected by rural employees.The main reason is the proposal for the coordinated development of the Beijing-Tianjin-Hebei Capital Economic Circle.It plays a role in boosting the population and labor loss in Hebei Province,and the intra-provincial mobility of labor.The number of rural workers has decreased dramatically.In addition,the actual situation of the other three cities is different,and the driving factors affecting the area planted with winter wheat are also different.