论文部分内容阅读
遥感技术已经成为实现地表信息提取的主要手段。以高分辨率影像为主要数据源,采用面向对象的多尺度分割算法,根据对象的光谱、形状等特征,实现了面向高分遥感数据的土地利用分类算法。该算法结合了面向地物对象和综合对象特征的分类方法,充分发挥了高分辨率影像进行精细地物分类的优势,得到了高精度的分类结果。通过西双版纳纳板河流域国家级自然保护区实例验证表明:该算法总体精度达到88.58%,Kappa系数达到0.77,精度符合应用要求,能够实现土地利用高精度、快速的分类。
Remote sensing technology has become the main means of extracting surface information. Using high-resolution image as the main data source, an object-oriented multi-scale segmentation algorithm was used to realize the land use classification algorithm for high-resolution remote sensing data according to the characteristics of the object’s spectrum and shape. This algorithm combines the classification methods of object-oriented objects and comprehensive object features, gives full play to the advantages of fine-grained object classification in high-resolution images and obtains high-precision classification results. The validation of the Xishuangbanna National Nature Reserve shows that the overall accuracy of the algorithm reaches 88.58% and the Kappa coefficient reaches 0.77. The accuracy accords with the application requirements, and the classification of land use with high accuracy and speed can be realized.