Wheat Planting Area Estimation Based On Satellite Remote Sensing Data And Mixed Effect Model

The current situation of my country’s arable land resources is not optimistic,and there are problems such as reduction of per capita arable land,poor arable land quality,and insufficient arable land resources.Under this background,accurately and timely grasping the production situation of my country’s main food crops is an important issue that needs to be solved urgently and is related to the national economy and people’s livelihood.It can make the agricultural policies and measures of the central and local governments more scientific,and provide the public with accurate food production statistics and reliable estimates for agricultural production.The survey of the sown area of crops in my country is relatively complicated.The sample survey of the sown area of the main varieties of crops is mainly based on the calculation of the overall provincial level,and it is difficult to meet the calculation of the subordinate data.At present,the application of spatial information 3S technology represented by remote sensing technology,geographic information system and global positioning system has created conditions for crop-to-ground sampling design and meeting the calculation of subpopulations by counties and villages,which will help us to obtain more accurate,smallscale satellite remote sensing data,but specific and accurate estimation methods need to be further studied.This paper uses satellite remote sensing data from 12 towns in Yanzhou District,Jining City,Shandong Province,uses pixel and area data to establish a mixed effect model,and uses small area estimation(SAE)to estimate crop planting area in a small area and classification mixed effect model prediction The method(CMMP)realizes the result of estimating the planting area of wheat,and the estimation accuracy is higher than that of directly using pixels;in addition,this article also considers the situation that the pixel data cannot be directly obtained,and will be based on The distance watershed segmentation algorithm is improved,and then applied to satellite remote sensing images to obtain pixel data of the wheat area,and then the same mixed effect model and two estimation methods are used to estimate the wheat planting area,and the Arcmap software and The combined application of python software,image segmentation algorithm and mixed effect model shows that its estimation accuracy is also higher than that when using pixel data directly.The main contribution of this paper is to use mixed effect model and SAE and CMMP methods on the basis of obtaining pixel and wheat area data.The existing data set contains the estimation target,and the existing data set does not contain the estimation target.All of them have achieved the effect of improving the prediction accuracy.In addition,the watershed algorithm has been improved to some extent,and the image segmentation algorithm is combined with the mixed effect model and the crop planting area prediction method.The effect of improving the estimation accuracy has also been achieved,which highlights the advantages of the method used in this article,and it is also rich The application content of remote sensing data in agriculture.