摘要
为了提高森林资源二类调查中小班遥感面积测量精度,本文利用博爱县国有林场部分遥感影像建立了目视解译标志,借助GIS进行了小班面积判读,并利用GPS实测了小班面积.遥感判读数据和实测数据的平均相对误差及相关系数结果表明,除了苗圃地和竹林地平均相对误差较大外,其余地类误差均较小,所有地类的遥感判读面积与实测面积均存在显著相关关系;依据90个小班数据所建立的各地类遥感判读面积与实测面积线性回归模型,均通过了显著性检验;利用30块小班数据对模型进行了卡方适合性检验,各地类回归模型均通过了检验,实测数据与模拟数据没有显著差别,能够在当地二类森林资源调查中实际运用.
In order to improve the accuracy of area measurement in the forest resource inventory,the paper use the part of remote sensing images of state-owned forest farms in Boai country and establish a visual interpretation,use GIS for area interpretation and use of GPS measured sub-compartment area.The average relative error and correlation coefficient results of interpretation of remote sensing data and real data showed that,in addition to the average relative error of the land types pf nursery and bamboo is more larger,the other land types are small,all in the class area and the measured area of remote sensing interpretation,there were significant correlation;Based on the data of 90,the paper established the linear regression model between measured area of remote sensing interpretation and real data,all passed the test of significance;using 30 sub-compartment data,the model of regression model around the class is fit according to the chi-square test,the measured data and simulation data are no significant difference between the two types,the model can be used to the local practical application of orest resources investigation.
出处
《河南科学》
2011年第1期44-48,共5页
Henan Science
基金
河南省科技攻关项目(0624050007)
河南农业大学林学院青年教师科研基金项目(20070001)
关键词
遥感图像
回归分析
判读模型
二类调查
the remote sensing image
regression analysis
interpreting model
forest management inventory
