基于地质统计学纹理的遥感影像分类研究进展

Advance in Remote Sensing Classification Based on Geostatistical Texture

传统的基于像素与像素基础上的遥感影像光谱分类方法忽视了邻近像素值之间潜在有用的空间信息,三十多年来,人们一直都在谋求利用遥感影像本身所固有的空间信息以加强光谱分类,尽管从事该方面研究的人一直都很少,其实现的手段主要依靠对原始影像的滤波,滤波的一般方式是生成纹理波段以指导接下来的分类。近年来,变异函数被用来表达空间依赖性,并取代简单的方差滤波成为了纹理分类的主要手段,在这篇综述性的论文中,笔者主要讨论了两类将基于地质统计学的纹理信息集成到遥感影像分类中的应用,它们代表了当前遥感影像纹理分类的主流。

Traditional spectral clarification of remotely sensed images relied on a pixel-by-pixel basis ignores the useful spatial information between the proximate pixels. Nevertheless, for some 30 years the spatial information inherent in remotely sensed images has employed to enhance spectral classification by a limited number of researchers. This has been achieved primarily by filtering the original imagery to derive texture ＇bands＇ for subsequent classification. Recently, the variogram has used to represent formally the spatial dependence in remotely sensed images and used in texture classification in place of simple variance filters, In this review, two methods of incorporating spatial information into the classification of remotely sensed images are discussed, which represent the main stream of current texture classification.