|
|
|
|
  • 软件名称:基于高光谱非线性向量空间的光谱曲线特征差异性分析
  • 软件大小: 0.00 B
  • 软件评级: ★★★
  • 开 发 商: 戴晓爱,杨晓霞,高孝杰,杨武年,贾虎军,杨叶,潘佩芬
  • 软件来源: 《遥感技术与应用》
  • 解压密码:www.gissky.net

资源简介

摘要: 地物光谱特征分析是对地物进行分类和匹配的基础,目前高光谱遥感技术应用在精细物种识别中主要采用波谱分析的方法。重点探索非线性空间里的相似性测度方法,由于光谱曲线表征复杂光谱成像的非线性过程,论文从空间目标的整体形状描述非空集合之间的差异,采用Fr′echet距离、Hausdorff 距离、Euclidian距离分别定义光谱特征曲线的距离,设计算法测量光谱向量之间的非线性相似程度。结果表明,采用Fr′echet距离、Hausdorff 距离、Euclidian距离度量光谱相似度的精度依次减弱,但依据Fr′echet距离的算法时间复杂度略高。基于Fréchet距离的方法充分考虑了曲线上点的位置信息及整体曲线的走势问题,其在精度、抗噪能力等方面均有提升,从而为分析光谱特征提供了可能的新途径。 关键词: 光谱相似度;  高光谱;  光谱曲线特征;  非线性向量空间     Abstract: It is the basis of classifying feature and matching that the characteristics of spectral analysis.Now the spectrum analysis is the main way in the fine species identification using the hyperspectral remote sensing technology.This paper focuses on exploring the nonlinear space in the inner similarity measure.Since the spectral curve of characterization can express the complex nonlinear spectral imaging process.The paper describes the non-empty set from the overall shape of space target.The distance between spectrum characteristic curve was defined Fr ′ echet distance,Hausdorff distance and Euclidean distance respectively and algorithm was designed to measure nonlinear similar degree between the spectral vector.The results show that the accuracy about Fr ′ echet distance,Hausdorff distance,the precision of the Euclidean distance to measure the spectral similarity is once weakened,but on the basis of Fr′echet distance algorithm time complexity is a little high.This method based on Fr′echet distance improves the accuracy and denoising performance,due to considering point position and trend of the curve,thereby which provides an effective tool to spectral analysis.

下载说明

·如果您发现该资源不能下载,请通知管理员.gissky@gmail.com

·为确保下载的资源能正常使用,请使用[WinRAR v3.8]或以上版本解压本站资源,缺省解压密码www.gissky.net ,如果是压缩文件为分卷多文件,请依次下载每一个文件,并按照顺序命名为1.rar,2.rar,3.rar...,然后鼠标右击1.rar解压.

·为了保证您快速的下载速度,我们推荐您使用[网际快车]等专业工具下载.

·站内提供的资源纯属学习交流之用,如侵犯您的版权请与我们联系.