Font Size: a A A

Information Extraction Of Citrus Orchard Based On Improved Support Vector Machine And Spatial Distribution Characteristic Analysis

Posted on:2020-11-18Degree:MasterType:Thesis
Country:ChinaCandidate:Y ZhongFull Text:PDF
GTID:2393330575999035Subject:Surveying and mapping engineering
Abstract/Summary:PDF Full Text Request
In recent years,the citrus industry has been influenced seriously by the occurrence of natural disasters such as Huanglong disease in Gannan,Therefore,it is necessary to conduct a research on citrus orchard,and obtain information of its area and spatial layout,so that which can provide some scientific ideas for produce management and making industrial policies..At present,the support vector machine(SVM)has been widely used in the classification of remote sensing information,but the information extraction for citrus orchards by SVM is rare,and the inappropriate parameter selection method will lead to the reduction classification accuracy of SVM.In this paper,the SVM parameter was optimized by the artificial bee colony algorithm(ABC)and constructed a ABC-SVM classification model,then extracted the citrus orchard information,which can achieve the analysis of the area dynamic changes and the spatial layout characteristics.(1)Construction and classification of the ABC-SVM model.The classification accuracy of SVM was taken as the objective function value of ABC algorithm.Then the ABC algorithm was used to optimize the penalty parameters and kernel function parameters of SVM.Combined ABC algorithm and SVM to construct the ABC-SVM classification model.After image preprocessing,selection of training samples and feature extraction,the ABC-SVM model was used to classify the image of the study area.The result showed that the classification accuracy of the ABC-SVM model was at least 85.74%,which was higher than SVM model.Therefore,the ABC-SVM model has a better effect and it is suitable for the extraction of citrus orchard information.(2)Dynamic change analysis of citrus orchard area.The change direction of citrus orchard area in the study period was quantitatively described by land use transfer matrix.The results showed that the citrus orchard area increased by 41.68 square kilometers from 2009 to2013,which verified the actual effect of the strategic measures of "mountain reconstruction" and "enriching the people with fruit" in Gannan.From 2013 to 2017,the total area of citrus orchards decreased by 172.31 square kilometers,reflecting the great impact of natural disasters such as huanglong disease on the citrus orchard industry.(3)spatial layout changes analysis of citrus orchards.Based on the area change of citrus orchard,constructed the planting concentration index and land use scale index,and analyzed the spatial layout changes of villages and towns,the results showed that the change of planting concentration index from 2009 to 2017 can be divided into the following four types:sustainable growth and falling type,stable type,and wave type.In addition,the change of land use scale index from 2009 to 2017 mainly includes the following three types: continuous falling type,stable type,and wave type.(4)correlation analysis of citrus orchard distribution.Took the villages and towns as the research unit,the global correlation and local correlation of citrus orchard distribution were analyzed through Moran's I index.The results showed that considered the concentration index as the variable,the overall study area showed very obvious spatial clustering characteristics from 2009 to 2017,and the clustering degree was first weakened and then increased.At the same time,the clustering types of citrus orchard planting concentration among the towns were mainly the H-H or L-H,and the towns belong to H-H type did not change and were relatively stable over time,while the towns belong to L-H type had significant changes.Considered the land use scale index as the variable,the overall study area showed very obvious spatial clustering characteristics and the degree of aggregation showed an increasing trend.At the same time,the clustering types of citrus orchard land use scale among the towns are mainly the H-H,L-L or L-H,and the towns belong to L-H type did not change and were relatively stable over time,while the towns belong to H-H and L-L type had significant changesBased on the above analysis,conclusions and suggestions can be drawn: citrus in Gannan has been commercialized.and it has formed the characteristics of large scale,continuous distribution and single crop cultivation,and which occupies a large proportion in the farmers and regional economy.Therefore,relevant departments must keep analyzing the correlation of spatial layout of citrus orchards,and take measures in disaster prevention,and reduce the losses of farmers.At the same time,we must pay attention to the reasonable planting of citrus,strengthen the construction of soil and water conservation measures in the garden,to promote the sustainable development of the ecological environment in Gannani.
Keywords/Search Tags:Artificial bee colony algorithm, support vector machine, area dynamic change, spatial layout change, spatia correlation
PDF Full Text Request
Related items