元胞自动机图像处理规则.doc

中图法分类号：TP79 文献标识码：A 文章编号： - - 文章索引信息: 海岸线遥感信息提取元胞自动机方法及其应用 冯永玖，（上海海洋大学海洋科学学院，上海 201306）摘要：从遥感资料中准确提取海岸线信息，是海岛海岸带遥感调查与海岸带综合管理的重要内容。为了对多类型海岸线进行准确，自下而上演化的元胞自动机遥感图像方向信息权重，构建了一种新的岸线提取算法，并在Matlab环境下开发实现包括海陆分离、离散地物去除和目标追踪的功能。以Landsat ETM遥感影像作为数据源，分别对人工海岸上海市长兴岛-横沙岛和淤泥质海岸上海市崇明东滩岸线进行，验证了岸线算法。研究表明，基于元胞自动机的提取方法，能够以较高精度人工海岸和淤泥质海岸的岸线信息。 关键词：海岸线提取；元胞自动机；方向信息；遥感影像A cellular automata approach to extract shoreline from remote sensing imageries and its application Feng Yongjiu, Han Zhen (College of Marine Sciences, Shanghai Ocean University, Shanghai 201306) Abstract: Extraction of shoreline information from remotely sensed imagery is an important issue for the remote sensing investigation of islands and comprehensive management of coastal areas. In order to accurately interpret the information of shorelines, a new algorithm was constructed by considering the direction information of imageries based on a “bottom-up” approach, cellular automata (CA). The algorithm was developed under the Matlab environment, which includes the functions such as mean smooth, land/water segmentation, discrete objects removal, and shoreline tracing. Using the proposed algorithm, the shorelines of artificially coast Changxing-Hengsha islands of Shanghai and muddy coast Chongming Dongtan of Shanghai were extracted from the Landsat ETM+ imageries. The results demonstrate that the CA based shoreline extractor can accurately detect shoreline information of different coast types (e.g. artificially coast and muddy coast) from remote sensing images. Key words: Shoreline extraction; Cellular automata; Direction information; Remote sensing imagery 引言 海岸线信息的提取研究，是海岛海岸带遥感调查的重要，关系到海岸带资源利用与环境管理。随着遥感资料的丰富和GIS技术的成熟，岸线信息越来越多实现了从遥感自动岸线信息的实质是遥感影像的边缘检测。对海岸线的遥感，通常采用以下方法实现：①微分检测方法，如Roberts、Prewitt、Sobel、Laplace、LOG及Canny等算子[]；②数学形态学检测方法[]；③智能检测方法[]等。研究认为Canny算子在人工海岸、基岩海岸、砂质海岸及粉砂淤泥质海岸等的岸线检测中均能够取得良好的效果[]，是海岸线检测的基本方法但是对于地物复杂、边缘特征或线特征破碎的遥感影像存在着明显的缺陷[]；而结合小波变换可以用来检测边缘，能够获取较好的轮廓效果[4]。通过非监督分类可以检测岸线的变化[11]。Li和Michiel从MSS/T