基于最小距离原理的自适性分类方法

Classification Method of Itself Adjustment Based on Minimum Distance Principle

在遥感图像的分类中,面对自然界所固有的随机性,时间和空间的变化以及遥感设备本身的误差和噪声等的影响,如何准确地把一个像素区分为某一地物类别,一直是遥感图像分类中的难点.首先介绍了最小距离原理,基此以比较的方式详细阐述了自适性最小距离分类方法的基本思路、步骤、特点以及可能存在的不足.

In the classification of the remote sensing image, it is always difficult to distinguish a category of image into some ground thing in the face of following factors： inborn randomness in nature, changes of time and places, errors existing in instrument, and the influence of noise. The minimum distance principle is introduced first. Then the basic thought, steps, characteristics and possible insufficiency of the classification method are stated in details by comparison.