| 基于波谱特征曲线法的遥感图像海岸线提取 |
| |
| 引用本文： | 尧雪娟.基于波谱特征曲线法的遥感图像海岸线提取[J].广东广播电视大学学报,2009,18(6):104-107. |
| |
| 作者姓名： | 尧雪娟 |
| |
| 作者单位： | 广东理工职业学院,广东广州,510091 |
| |
| 摘 要： | 本文利用遥感图像的反射波谱信息，提出了海岸线提取的一种方法。首先，根据遥感图像中水体独特的反射波谱特征，提出了以特征曲线法来进行水陆分离，其核心是基于地物波谱的相似度，以水体反射波谱曲线作为中心判据，当目标像元的特征曲线符合中心判据时，将目标像元分类为水体，反之为非水体，从而实现水陆分离。中心判据由水体波谱在遥感图像各波段的分布特征来确定。然后利用形态学方法，对水陆分离后的图像进行海岸线提取。本文以Quick Bird图像和IKONOS图像为例，利用ENVI软件编程，实现了对归一化后的遥感图像的海岸线提取。结果表明所获取的海岸线连续性好、信噪比高，算法简单且适用性强。进一步研究自适应获取中心判据的方法，可使该算法用于工程化运算。
|
| 关 键 词： | 遥感图像 海岸线 波谱特征曲线 水陆分离 形态学 |
Coastline Extraction of Remote Sensing Images Based on Spectrum Character Curve Method |
| |
| YAO Xue-juan.Coastline Extraction of Remote Sensing Images Based on Spectrum Character Curve Method[J].Journal of Guangdong Radio & Television University,2009,18(6):104-107. |
| |
| Authors: | YAO Xue-juan |
| |
| Institution: | YAO Xue-juan (Guangdong Polytechnic Institute, Guangzhou, Guangdong, China, 510091) |
| |
| Abstract: | By using the spectrum of remote sensing images, an algorithm to detect coastline is proposed in this paper. Firstly, according to the unique spectrum characteristic of waters in remote sensing images, we propose a spectrum-character-curve method to segment remote sensing image into waters-to-land image. The spectrum-character-curve method is mainly based on spectrum similarity; and the waters spectrum curve is taken as the central criterion; when spectrum curve of target pixel accords with the criterion, the target pixel is classified to waters, otherwise, it belongs to land, so two-value segmenting of waters and land is realized. And the central criterion is determined by waters＇ spectrum distribution in each band of remote sensing images. Then, with two-value morphology method, the coastline is extracted from waters-to-land image. Taking QuickBird images and TM images as examples, coastline extraction experiments are executed of normalized images by ENVI Programming. The results indicate the detected coastline is continuous and with high signal-to-noise ratio （SNR）, also this algorithm is simple and reliable. This algorithm can be applied to engineering computation while the adaptive selection method of central criterion is further realized. |
| |
| Keywords: | remote sensing image coastline spectrum character curve segmenting of waters and land morphology |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
