简介：Weusedgeographicinformationsystemapplicationsandstatisticalanalysestoclassifyyoung,prematureforestareasinsoutheasternGeorgiausingcombineddatafromLandsatTM5satelliteimageryandgroundinventorydata.Wedefinedprematurestandsasforestswithtreesupto15yearsold.Weestimatedtheprematureforestareasusingthreemethods:maximumlikelihoodclassification(MLC),regressionanalysis,andk-nearestneighbor(kNN)modeling.Overallaccuracy(OA)ofclassifyingtheprematureforestusingMLCwas82%andtheKappacoefficientofagreementwas0.63,whichwasthehighestamongthemethodsthatwehavetested.ThekNNapproachrankedsecondinaccuracywithOAof61%andaKappacoefficientofagreementof0.22.RegressionanalysisyieldedanOAof57%andaKappacoefficientof0.14.WeconcludethatLandsatimagerycanbeeffectivelyusedforestimatingprematureforestareasincombinationwithimageprocessingclassifierssuchasMLC.
客服QQ：30444492琼网文【2021】1550-113号
增值电信业务经营许可证：琼B2-20210322
出版物经营许可证：新出发龙华出字第(2021)009号
广播电视节目制作经营许可证：(琼)字第00779号
版权所有 ©2002-2024 期刊网 琼ICP备2021005105号
