多特征融合的光学影像围填海变化检测与地物识别

发布时间：2018-03-24 17:35

本文选题：高空间分辨率　切入点：围填海　出处：《河南理工大学》2016年硕士论文

【摘要】：围填海作为人类开发利用海洋资源的主要方式,是拓展建设和社会发展空间的重要途径。适度有序的围填海,对于减轻耕地保护压力,有效的拉动投资,调整海洋产业布局和城市建设布局,治理、修复海域、海岸带和海洋的生态环境,具有重要的意义。但过度无序的围填海会极大地损害国家利益,造成相当多的生态环境问题。遥感技术以其较低的成本,快速、准确、大范围的对地观测能力被引入到海域使用情况调查中,成为海域监测的有力手段。本文利用高空间分辨率光学影像数据对围填海的动态进行监测研究,及时获取围填海的动态信息,为有效地保障国家权益提供技术支撑。本文研究了如何利用高分影像(本文中出现的高分影像如果没有特指,一般指的是高空间分辨率光学遥感影像)快速获取围填海的变化信息、自动化的识别出变化区域地物类别,提出一套有效的解决方案,主要研究工作如下:(1)分析当前利用遥感影像围填海变化检测方法,进行实验分析,发现现有遥感影像用于围填海变化检测存在的一些问题。把K-mean聚类分析的思想引入的围填海变化检测中,该方法能够解决变化斑块区域在变化检测强度图上灰度空间上尺度不一致的难题,并进行了试验予以验证。(2)提出了一种多特征融合的围填海变化斑块识别的算法,对变化斑块进行类别识别,总体识别精度达到了80.47%。
[Abstract]:As the main way for human beings to develop and utilize marine resources, reclamation is an important way to expand the space for construction and social development. It is of great significance to adjust the layout of marine industry and urban construction, to manage and repair the ecological environment of the sea areas, coastal zones and oceans. Remote sensing technology, with its low cost, rapid, accurate and extensive Earth observation capability, has been introduced into the survey of sea area use. In this paper, high-resolution optical image data are used to monitor and study the dynamic reclamations in the sea area, and the dynamic information of the reclamation is obtained in time. In order to provide technical support for the effective protection of national rights and interests, this paper studies how to make use of high-score images. Generally refers to the high-spatial resolution optical remote sensing image) to quickly obtain the change information of the reclamation area, automatically identify the category of the change area, and put forward a set of effective solutions. The main research work is as follows: (1) analyzing the current detection methods of reclamation changes using remote sensing images, and carrying out experimental analysis. Some problems existed in the existing remote sensing images used in reclamation change detection are found. The idea of K-mean clustering analysis is introduced into the detection of reclamation change. This method can solve the problem of inconsistent scales in gray scale space in the change detection intensity map of patch region, and is verified by experiments. (2) A multi-feature fusion algorithm is proposed to identify the patch changes in the reclaimed sea. The recognition accuracy of the change plaque is 80.47.
【学位授予单位】：河南理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P237;P756.8

【相似文献】