【作者】 陈忠；

【导师】 赵忠明；

【作者基本信息】 中国科学院研究生院（遥感应用研究所） ， 地图学与地理信息系统， 2006， 博士

【摘要】 遥感技术的发展，使我们能够获得极其丰富的信息，尤其是近年来高分辨遥感图像的出现更扩大了对自然界观察的视野。但是面临的挑战是如何处理和应用这些数据，使之能转换为急需被应用的信息。IKONOST和QuikBird等高分辨遥感图像表现出地物更多的信息诸如光谱、形状、纹理以及上下文等。尽管卫星遥感数据分类技术有了长足的发展，但是对于高分辨遥感图像来说，利用单一传统的分类方法不仅会导致分类精度降低，而且也会造成空间数据大量冗余、资源浪费。 因此，本文紧紧围绕提高高分辨遥感图像的分类精度这一中心环节，以IKONOS高分辨率遥感图像为主要信息源，重点从分类方法和影响分类精度的尺度效应两个方面加以论述。在对前人工作研究的基础上，提出了针对高分辨遥感影像的新的分类算法和一种分类最佳尺度的确定方法。 本文主要研究成果如下： 1) 在对传统分类算法进行研究的基础上，利用混合判别规则来代替传统的单一判别规则的多分类器融合技术来提高高分辨率遥感图像分类精度。并把支持向量机(SVM)引入到高分辨遥感图像分类中。 2) 研究了面向对象高分辨遥感图像分类技术。并结合高分辨遥感图像所表现出来的光谱和形状特征，提出了3种适合高分辨遥感图像的多尺度、多特征图像分割算法。 3) 利用初始分割得到的图像对象，通过监督模糊分类算法来对各个区域对象进行多特征模糊分类。 4) 研究了面向对象高分辨遥感图像分类精度和面积尺度的关系。提出了一种二维面积直方图的统计方法，并用来计算各类别最佳面积尺度。更多还原

【Abstract】 The development of the remote sensing technology makes us obtain very abundant information of nature, especially with the appearance of high resolution remote sensing image it extends the visual field of the nature. But the challenge that faces us is how to make use of the data effectively and obtain more useful information through some processing. High resolution remote sensing data such as QuikBird and IKONOS have a lot of characteristics such as spectral, shape, texture and context and so on compared to the other remote sensing data. Though the technology of the remote sensing image classification has made considerable progress, it will result in not only reducing the classification accuracy but also making the spatial data redundant and wasting the resource when the single traditional classification method is applied to the high resolution remote sensing image.So, with the IKONOS data as the main source, this thesis discusses the classification algorithms and scale-effect which affect the classification accuracy in order to improve the high resolution remote sensing image classification accuracy. The new classification algorithms suitable for the high resolution remote sensing image are brought forward and the relationship between the classification accuracy and the scale effect is discussed.This dissertation includes the following research products and originalities:1) Based on the research of the traditional classification, brings forward a multiple classifier system with a mixed combining rule for decision not a single rule to improve the classification accuracy. Apply the support vector machine to the high resolution remote sensing image classification.2) Study the object-oriented classification technology of high resolution remote sensing image. Combined with the information obtained from the high resolution remote sensing image, three multi-scale, multi-characteristic image segmentation algorithms were brought forward.3) Bring forward a supervised fuzzy classification algorithm with multiple更多还原