光学遥感影像土地利用分类方法综述

土地是实现经济社会高质量发展的基础性资源，光学遥感影像以时间、空间、光谱等特有的优势为土地利用变化的发现、识别、分析提供了很好的支持，利用光学遥感技术提取土地利用变化信息已经成为重要的研究领域。论文以近年来土地利用分类相关文献为基础，分别从数据源、处理流程、处理精度、计算效率等角度对土地利用分类应用较为广泛的目视解译、监督分类、非监督分类、深度学习等四类方法进行了分析和总结。最后展望了光学遥感影像土地利用分类方法的发展方向

Review of Land Use Classification Methods Based on Optical Remote Sensing Images

Land is used to realize the economic and social development of high quality basic resources.Such as time, space, spectrum is used as optical remote sensing image unique advantages for the discovery of land use change, the identification, analysis provides a good support, optical remote sensing technology is utilized to extract information of land use change has become an important area of research. Based on the literature on land use classification in recent years, four widely used land use classification methods, namely visual interpretation, supervised classification, unsupervised classification and deep learning, were analyzed and summarized from the perspectives of data sources, processing process, processing accuracy and computational efficiency. Finally, the development direction of land use classification methods for optical remote sensing images is prospected.