论文部分内容阅读
土地利用/覆被专题信息的快速、高效、准确提取是遥感图像处理研究的重要方向。传统的遥感分类方法常依靠像元的光谱值,未充分利用影像的空间信息。本文将面向对象影像分割和支持向量机方法相结合,复合光谱和纹理信息,建立了Object-SVM分类模型,并与面向对象的模糊函数和基于像元的SVM方法相比较,探寻区域尺度土地利用/覆被信息提取方法。结果显示,Object-SVM模型有效地提高了遥感图像的分类精度和分类效率,对于区域尺度影像的快速、准确、客观的信息提取具有实际意义。
The rapid, efficient and accurate extraction of land use / cover thematic information is an important direction of remote sensing image processing. Traditional remote sensing classification often relies on the spectral values of pixels and does not make full use of spatial information of images. In this paper, object-SVM classification model is established by combining object-oriented image segmentation with support vector machine, compound spectrum and texture information. Compared with object-oriented fuzzy functions and pixel-based SVM methods, this paper explores the application of regional scale land use / Cover information extraction method. The results show that the Object-SVM model can effectively improve the classification accuracy and classification efficiency of remote sensing images, which has practical significance for fast, accurate and objective information extraction of regional scale images.