[email protected] | |
3275638434 | |
Paper Publishing WeChat |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Research on the Classification of High Resolution Image Based on Object-Oriented and Class Rule
Li Chaokui, Fang Jun, Wu Baiyan and Chen Jianhui
Full-Text PDF XML 1348 Views
DOI:10.17265/2159-581X/2018.01.003
With the development of remote sensing technology, the spatial resolution, spectral resolution and time resolution of remote sensing data are greatly improved. How to efficiently process and interpret the massive high resolution remote sensing image data for ground objects, which are of spatial geometry and texture information, has become the focus and key issue in the field of remote sensing research. A new method of the classification (OCRC (object-oriented and class rule classification)) of remote sensing, which is of object-oriented and rule, has been presented in this paper, that is, through the discovery and mining the knowledge of spectrum and spatial characteristics of high-resolution remote sensing image, establish a multi-level network image object segmentation and classification structure of remote sensing image to achieve classification and accuracy assessment accurately and quickly for ground targets. Selected the worldview-2 image data in the Zangnan area as a study object, using the OCRC to verify the experiment which was a combination of the mean variance method, the maximum area method and the accuracy comparison to analysis selected three kinds of optimal segmentation scale and established a multi-level image object network hierarchy for image classification experiments. The results show that the OCRC can enable the high resolution image classification results similar to the visual interpretation of the results, and has higher classification accuracy. The overall accuracy and Kappa coefficient of the object-oriented rule classification method are 97.38%, 0.9673; compared with object-oriented SVM method, respectively higher than 6.23%, 0.078; compared with object-oriented KNN method, respectively more than 7.96%, 0.0996. The extraction precision and user accuracy of the buildings compared with object-oriented SVM method, respectively higher than 18.39%, 3.98%, respectively better than the object-oriented KNN (K-Nearest Neighbor) method 21.27%, 14.97%.
Object-oriented, rules, high resolution, multi-scale segmentation.