面向地理国情监测的遥感影像变化检测方法研究

发布时间：2018-05-29 14:52

本文选题：面向对象 + 多级分解　；参考：《辽宁工程技术大学》2013年硕士论文

【摘要】：面向地理国情监测重大工程应用,本文重点研究高分辨率遥感影像变化检测方法及技术,建立变化检测工艺流程,研制变化检测软件系统,实现方法及系统的初步应用,为地理要素变化检测提供技术手段与软件平台。主要包括以下几个方面的内容： (1)基于多级分解模型的变化检测方法研究。分析典型地物地类特征,建立多级分解模型,对前后两期遥感影像使用分解统一模版,以最小分解子块作为变化检测基本处理单元,探索多级分解的变化检测分解思路。 (2)基于模糊规则的变化信息预测。设定判断阈值,将整幅影像归类为：确定变化、确定未变化和模糊变化三个区域；对模糊变化区域依照模糊判断规则归类,辅助人工判读,获得变化区域范围信息。 (3)面向对象的变化信息提取。通过自适应阈值分析,模糊判断获得的变化检测结果为发生变化与未发生变化的二值掩膜影像。选择SVM或GLC分类器对变化区域进行分类,得到准确的变化信息。 (4)面向地理国情监测的变化检测工艺流程。在测绘科学研究院研发的地理国情地表覆盖解译系统的基础上,将本文所提出的多级分解模糊判断变化检测方法工艺流程进行集成,优化基于地理国情监测的高分辨率遥感影像地表覆盖解译系统。
[Abstract]:In this paper, the methods and techniques of high-resolution remote sensing image change detection are studied, the process flow of change detection is established, the software system of change detection is developed, and the method and the preliminary application of the system are studied. It provides technical means and software platform for the change detection of geographical elements. It mainly includes the following aspects: 1) Research on change detection method based on multilevel decomposition model. This paper analyzes the characteristics of typical ground objects, establishes a multilevel decomposition model, uses a unified decomposition template for the two periods of remote sensing images, and takes the minimum decomposition sub-block as the basic processing unit for change detection, and probes into the idea of multi-level decomposition for the change detection and decomposition. 2) Prediction of change information based on fuzzy rules. The whole image is classified into three regions: determine the change, determine the unchanged and fuzzy change, and classify the fuzzy change area according to the fuzzy judgment rule, and then obtain the range information of the change region by manual interpretation. 3) object-oriented change information extraction. Through adaptive threshold analysis, the change detection results obtained by fuzzy judgment are binary mask images with or without change. Select SVM or GLC classifier to classify the change region and get accurate change information. 4) change detection process for geographical situation monitoring. On the basis of the ground cover interpretation system developed by the Institute of surveying and Mapping Sciences, the process flow of the multi-level decomposition fuzzy judgment change detection method proposed in this paper is integrated. To optimize the land cover interpretation system of high resolution remote sensing image based on geographical situation monitoring.
【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

【参考文献】