Skip to Main Content
Object-oriented classification of remote sensing image is that the image is divided into a series of image objects based on objects' spectrum, shape and texture features, adopting the technology of fuzzy classification to achieve classification and information extraction. In this paper, SPOT5 image of a hypaethral mine as an study area, object-oriented classification was used to study the mine information extraction and classification. After completing multi-scale image segmentation and establishment of classification rules, object-oriented classification of experimental area image was accomplished. The results show that the classification accuracy based on object-oriented technology of image classification is higher, by compared with the results based on traditional technology of image classification. The study in this paper demonstrates the superiority of object-oriented technology of image classification and it's particularly suitable for information extraction of high-resolution remote sensing data.