基于TM遥感影像的诸暨市森林资源监测

林业资源管理 ›› 2011, Vol. 0 ›› Issue (6): 104-109.

基于TM遥感影像的诸暨市森林资源监测

陈军, 邱保印

1.浙江省森林资源监测中心,杭州 310020;
2.浙江农林大学经济管理学院,浙江临安 311300

收稿日期:2011-09-09 修回日期:2011-10-21 出版日期:2011-12-28 发布日期:2020-12-18
作者简介:陈军(1979-),男,浙江浦江人,工程师,在读硕士,主要从事林业规划设计工作。

Forest Resources Monitoring for Zhuji CityBased on Thematic Mapper(TM) Imagery

CHEN Jun¹, QIU Baoyin²

1. Zhejiang Forestry Resources Monitoring Center,Hangzhou 310020,Zhejiang,China;
2. School of Economics and Management,Zhejiang A &F University,Lin'an 311300,Zhejiang,China

Received:2011-09-09 Revised:2011-10-21 Online:2011-12-28 Published:2020-12-18

摘要/Abstract

摘要： 借助遥感技术提高国家森林资源连续清查效率具有重要的意义。利用最大似然法、BP 神经网络和最近邻算法3种不同的分类方法对诸暨市森林资源进行监测,并将分类结果与二类森林资源调查数据作对比。结果表明,3种分类方法都能较高精度地监测诸暨市不同森林类型总面积,精度在77.53%~83.18%之间;但是在乡镇尺度上,3种分类方法的精度都不理想,总相对均方根误差分别为41.83%,44.91%和44.18%。除灌木林外,以上3种分类方法在精度上没有显著差异(P>0.5)。今后研究应从多源遥感影像融合等技术上提高像元尺度上的分类精度。

关键词: 最大似然法, BP 神经网络, 最近邻算法, 森林资源监测, 诸暨市

Abstract: It has a significant meaning to increase the efficiency of the national forest inventory using remote sensing technology.Maximum likelihood,back propagation neural network,and k-nearest neighbor were applied to monitor forest resources for the the Zhuji city.Classification results were compared with forest inventory data.Results showed that these three methods accurately estimated the total area of different forest types with accuracy between 77.53% and 83.18%.However,the accuracies of these three methods are low at town level with relative root mean square error(RMSEr) of 41.83%,44.91%,and 44.18% respectively.Except for shrub forest,these three classification methods are not significantly different(p>0.5).Some techniques,such as multi-source remote sensing image fusion,should be used to increase classification accuracy in future study.

Key words: maximum likelihood, back propagation neural network, k-nearest neighbor, forest resources monitoring, Zhuji City

中图分类号:

S771.8

陈军, 邱保印. 基于TM遥感影像的诸暨市森林资源监测[J]. 林业资源管理, 2011, 0(6): 104-109.

CHEN Jun, QIU Baoyin. Forest Resources Monitoring for Zhuji CityBased on Thematic Mapper(TM) Imagery[J]. FOREST RESOURCES WANAGEMENT, 2011, 0(6): 104-109.

参考文献

[1] 王六如,李崇贵.森林资源清查移动GIS 系统研制[J].林业科学,2010,46(8):174-180.
[2] 林辉,熊育久,孙华,等.湖南省森林资源连续清查遥感应用研究[J].中南林业科技大学学报,2007,27(4):27-32.
[3] 王连宽,李晔,王聪.森林资源连续清查体系存在的问题及对策措施[J].内蒙古林业调查设计,2007,30(1):24-31.
[4] 陈雪峰,曾伟生,熊泽彬,等.国家森林资源连续清查的新进展—关于国家森林资源连续清查技术规定的修订[J].林业资源管理,2004(5):40-45.
[5] 苏顺,邵保华.“3S”技术在森林资源连续清查中的应用[J].林业科技,2005,30(4):23-24.
[6] 余松柏,林中大,叶金盛.GPS在森林资源清查中的应用[J].南京林业大学学报:自然科学版,2004,28(1):84-86.
[7] 党永峰.遥感技术在森林资源连续清查中的应用——以利用遥感技术分析森林植被地类的动态变化为例[J].林业资源管理,2004(6):63-65.
[8] 刘尚斌,孙涛,黄国胜.遥感技术在森林资源连续清查中的应用研究[J].林业资源管理,2000(6):51-57.
[9] 王立海,赵正勇.基于BP 神经网络的针阔混交林TM遥感图像自动分类技术研究[J].林业科学,2005,41(6):94-100.
[10] Verbeke L P C,Vabcoillie F MB,Dewulf R R.Reusing back-propagating artificial neural network for land cover classification in tropical savannahs[J].International Journal Remote Sensing,2004,25:2747-2771.
[11] 施拥军,徐小军,杜华强,等.基于BP神经网络的竹林遥感监测研究[J].浙江林学院学报,2008,25(4):417-421.
[12] 李祚泳.用BP 神经网络实现多波段遥感图像的监督分类[J].红外与毫米波学报,1998,17(2):153-156.
[13] 于秀兰,钱国蕙,周建林,等.多光谱遥感图像BP网分类器学习样本选取法的研究[J].红外与毫米波学报,1999,18(6):449-454.
[14] 张孝飞,黄河燕.一种采用聚类技术改进的KNN文本分类方法[J].模式识别与人工智能,2009,22(6):936-940.
[15] Thessler S,Sesnie S,Ramos Bendaña Z S,et al.Using k-nn and discriminant analyses to classify rain forest types in a Landsat TM image over northern Costa Rica[J].Remote Sensing of Environment,2008,112(5):2485-2494.
[16] Gjertaen A K.Accuracy of forest mapping based on Landsat TM data and a kNN-based method[J].Remote Sensing of Environment,2007,110(4):420-430.
[17] Haapanen R,Ek A R,Bauer M E,et al.Delination of forest/nonforest land use classes using nearest neighbor methods[J].Remote Sensing of Environment,2004,89(3):265-271.
[18] Mcroberts R E.Using satellite imagery and the k-nearest neighbors technique as bridge between strategic and management forest inventories[J].Remote Sensing of Environment,2008,112(5):2212-2221.
[19] Gu D,Gillespie A.Topographic normalization of Landsat TM images of forest based on subpixel sun-canopy-sensor geometry[J].Remote Sensing of Environment,1998,64(2):166-175.
[20] 徐小军,杜华强,周国模,等.Erf-BP混合像元分解及在森林遥感信息提取中应用[J].林业科学,2011,47(2):30-38.
[21] Hudak A T,Crookston N L,Evans J S,et al.Nearest neighbor imputation of species-level,plot-scale forest structure attributes from LiDAR data[J].Remote Sensing of Environment,2008,112(5):2232-2245.
[22] Fuchs H,Magdon P,Klein C,et al.Estimating aboveground carbon in a catchment of the Siberian forest tundra:Combining satellite imagery and field inventory[J].Remote Sensing of Environment,2009,113(3):518-531.