传统的基于像素的遥感影像处理方法都是基于遥感影像光谱信息极其丰富，地物间光谱差异较为明显的基础上进行的。对于只含有较少波段的高分辨率遥感影像，传统的分类方法，就会造成分类精度降低，空间数据的大量冗余，并且其分类结果常常是椒盐图像，不利于进行空间分析。 为解决这一传统难题，模糊分类技术应运而生。模糊分类是一种图像分类技术，它是把任意范围的特征值转换为0到1之间的模糊值，这个模糊值表明了隶属于一个指定类的程度。通过把特征值翻译为模糊值，即使对于不同的范围和维数的特征值组合，模糊分类能够标准化特征值。模糊分类也提供了一个清晰的和可调整的特征描述。 本文采用面向对象的影像分类方法。考虑了对象的不同特征值，例如光谱值，形状和纹理。结合上下文关系和语义的信息，这种分类技术不仅能够使用影像属性，而且能够利用不同影像对象之间的空间关系。在对诸多对象进行分类后，在进行精度分析。在此研究提出了一种面向对象的方法结合模糊理论把许多的对象块分成不同的类别。这一过程主要有两个步骤：第一个步骤是分割。图像分割将整个图像分割成若干个对象，在对所有对象进行分类，在这个过程中，分割尺度的选择会影响到后续的分类结果和精度；第二个步骤是分类。在这个步骤中，特征值的选择和隶属度函数的选择都对分类结果有着至关重要的影响。完成分类后，可以进行土地变换检测等后续工作。

The traditional pixel-based remote sensing image processing approaches exploit the features of abundant spatial information and explicit difference of spectrum between different objects. Applying the traditional classifying algorithms to the high resolution remote sensing image with few spectral bands will cause the low classification rate, redundancy spatial data, and output image with pepper-salt noise. Fuzzy classification is a technique that basically translates feature value of arbitrary range into fuzzy values between 0 and 1, indicating the degree of membership to a specific class. By translating feature values into fuzzy values, fuzzy classification can standardize features and allows the combination of features, even of very different range and dimension. Fuzzy classification also provides a transparent and adaptable feature description. Object-based image classification, which is based on fuzzy logic, allows the integration of a broad spectrum of different object features such as spectral values, shape, and texture. Such classification techniques, incorporating contextual and semantic information, can be performed using not only image object attributes, but also the relationship among different image objects.This research has main two steps. First step is segmentation. During this process, how to choose fine scale is very important. Second step is classification. In this step, two factors are very essential for classification. Creating feature space is one of the factors, another is fuzzy logic. After these two steps, we can extract some classes that we are interested in.