论文部分内容阅读
分形几何学正在变为一种模拟自然现象的有用工具。作为对欧几里德概念的一个补充,分形便于对自然边界和表面中复杂性做出更精确的表述。本文的目的是以C语言介绍和实现遥感资料分形测量的三种算法。这三种方法是:分线法(等值线法),方差图法和三角棱柱方法。分别利用这三种方法来计算由美国国家航空和宇航局标定的航空多光谱扫描器获取的遥感资料的分维数。这些资料于1990年1月获得,是西波多黎各的一部分,像元空间分辨率为30米。本文说明这三种方法、它们在PC兼容机环境中的实现以及在遥感图像数据中应用这些算法的一些结果。
Fractal geometry is becoming a useful tool to simulate natural phenomena. As a complement to the Euclidean concept, fractals facilitate a more accurate representation of the natural boundaries and superficial complexities. The purpose of this paper is to introduce and implement three algorithms for fractal measurement of remote sensing data in C language. The three methods are: the line method (contour method), the variance method and the triangular prism method. The three methods are used to calculate the fractal dimension of remote sensing data acquired by the NASA-calibrated multispectral scanner. This information was acquired in January 1990 and is part of Western Puerto Rico, with a pixel spatial resolution of 30 meters. This article describes these three methods, their implementation in a PC-compatible environment and some of the results of applying these algorithms in remote sensing image data.