极化相干最优与角度信息相融合的道路提取方法研究

发布时间：2018-02-02 14:41

本文关键词： 道路提取极化相干最优 SAR影像角度信息提取道路轨迹模型　出处：《辽宁工程技术大学》2013年硕士论文　论文类型：学位论文

【摘要】：道路是现代交通体系的主体,对国家经济和基础地理信息建设有着重大意义。在光学遥感中,道路提取是经典的研究课题,取得了较多的研究成果。但直接应用这些方法从SAR影像中提取道路信息往往效果不佳。因而,SAR影像道路提取成为影像信息提取中重要的研究内容。本文根据SAR影像中道路的形状、灰度和极化相干等特征,提出了一种极化相干最优与角度信息相融合的SAR影像道路提取方法。该方法将极化相干最优分类的道路类信息与角度信息提取的灰度信息相结合作为道路的局部特征,引入道路轨迹模型提取道路的全局特征,最后综合这两种特征实现SAR影像道路提取。本文的工作和创新主要包括以下几个方面： (1)根据极化干涉SAR的相干最优原理,对SAR影像进行极化相干最优处理,统计分析了几种典型地物的极化相干最优系数的分布情况。并将极化相干最优分类与几种典型分类方法进行比较,确定了本文采用极化相干最优分类的方法提取道路类信息。 (2)选取道路中心点,用不同的数学模型进行拟合实验,通过比较分析确定了傅里叶函数作为道路轨迹模型。 (3)将极化相干最优分类后的道路类信息与角度信息提取中的灰度信息相融合,作为道路提取过程中矩形模板前进方向的判断依据。 (4)定义了初始化策略、前进策略、旋转策略和停止策略,以Matlab软件为平台,实现了SAR影像的道路提取。
[Abstract]:The road is the main body of modern traffic system, is of great significance to the national economy and the construction of fundamental geographic information. In the optical remote sensing, road extraction is a research topic of classic, many results have been achieved. But the direct application of these methods in road extraction from SAR image information is often ineffective. Therefore, SAR image road extraction as an important research content of image information extraction.
According to the SAR image in the shape of the road, intensity and polarization coherent characteristics, proposed a SAR image road extraction and optimal polarization coherent angle information fusion method. The method of local feature extraction of road information and gray information of angle information of polarization coherent optimal classification the combination of road, global features the introduction of the trajectory model of road extraction, finally integrate the two kinds of feature extraction for SAR image.
The work and innovation of this article mainly include the following aspects:
(1) according to the principle of optimal coherent polarimetric SAR interferometry, the image of SAR optimal polarization coherent, statistical analysis of the distribution of some typical features of the optimal polarization coherence coefficient. And the optimal polarization coherent classification and several typical classification methods were compared to determine the optimal polarization coherent method used in this paper to extract road information classification.
(2) select the center of the road, use different mathematical models to fit the experiment, and determine the Fu Liye function as the road trajectory model by comparison and analysis.
(3) we integrate the information of the best road information and the gray information in the extraction of angle information.
(4) the initialization strategy, the forward strategy, the rotation strategy and the stopping strategy are defined, and the road extraction of SAR images is realized on the platform of Matlab software.

【学位授予单位】：辽宁工程技术大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P237;P225.1

【参考文献】