Affiliations: [a] Chongqing Key Laboratory of Spatial Data Mining and Big Data Integration for Ecology and Environment, Rongzhi College of Chongqing Technology and Business University, PR China | [b] College of Land and Resources, China West Normal University, Sichuan, China
Correspondence:
[*]
Corresponding author. Zhongsheng Chen, College of Land and Resources, China West Normal University, Nanchong 637009, Sichuan, China. E-mail: chenzhs@cwnu.edu.cn.
Note: [1] E-mail: chen_gb1982@163.com
Abstract: With the rapid development of the economy, the demand for urban land resources is also growing. How to make more rational use of land resources and make more rational planning of cities become a major problem in current economic development. At present, the use of remote sensing images to classify urban land use areas has become a research hot spot. However, the traditional classification accuracy rate using the maximum likelihood classification method needs to be improved. How to improve the classification accuracy rate of urban land use area of remote sensing image has become the focus and key of the research. Both rough sets and fuzzy sets are mathematical methods for dealing with uncertain problems. The rough fuzzy sets generated by the combination of the two can solve the problem of information loss due to the rough set discretization process. Based on the advantages of fuzzy rough sets, this paper applies fuzzy rough sets to the study of urban land use area classification of remote sensing images, so as to improve the accuracy of urban land use area classification of remote sensing images. Firstly, the spectral features and texture features of the remote sensing image are extracted after preprocessing the remote sensing image. Secondly, using the domain relationship fuzzy rough set reduces the extracted features. Finally, the support vector machine is used to classify the reduced feature set, and the classification of urban land use area is realized. In the simulation experiment, the classification accuracy is evaluated by the overall classification accuracy, Kappa coefficient, and single class classification success index. The evaluation data shows that the fuzzy rough set is applied to the remote sensing image urban land use area classification, which has a good application effect.
Keywords: Urban land use area, remote sensing image, fuzzy rough set, support vector
machine
