摘要
半变异函数在获取遥感影像地物最优尺度的研究中发挥着极其重要的作用。为了获取更高精度的半变异函数曲线模型的变程属性,首先定义5个影响变程精度的参数,并对其进行初始化;同时定义两个衡量参数优化结果的指标;然后利用控制变量的方法,对上述参数进行逐个优化,并利用优化后的参数获取变程值;最后,本文提出最优观测尺度的选择方法,在最优变程值的基础上,通过对影像空间结构的定量描述和Shannon采样定理,得到最优观测尺度。研究结果表明,积分变程能够较好的实现半变异函数对空间结构的定量描述,研究区遥感影像分辨率≤40 m时,能够满足观测需要。
The semivariogram has played an important role in the study of the optimal scale of Remote Sensing imagery.In order to obtain more accurate range of Semivariogram curve,we first define and initialize five parameters,which may affect its attribute values.Meanwhile,two indicators are defined to measure the optimization results of the parameters.Then the method of controlling variables is used to optimize the parameters one by one,and the optimized parameters are used to obtain the range.At the end of the paper,a method of selecting the optimal observation scale is proposed.Based on the optimal range,the optimal observation scale is obtained by quantitative description of image spatial structure and Shannon sampling theorem.The results show that Integral variation can better realize the quantitative description of spatial structure by Semivariogram.And when the resolution of the image is less than 40m,it can meet the needs of observation.
作者
骆秀齐
郝庆雨
LUO Xiuqi;HAO Qingyu(College of Geoscience and Surveying Engineering,China University of Mining and Technology(Beijing),Beijing 100083,China)
出处
《北京测绘》
2019年第3期274-279,共6页
Beijing Surveying and Mapping
关键词
遥感
半变异函数
参数优化
变程
定量描述
最优尺度
remote sensing
semivariogram
parameter optimization
range
quantitative description
optimal scale
