土地覆盖遥感制图方法研究现状与展望

区域尺度土地利用/覆盖分类研究是当今国际上开展土地利用/覆盖变化研究的重要领域之一。及时、准确地获取地球表面特性对于掌握人类和自然现象之间的关系和相互作用至关重要。本文根据土地覆盖遥感分类方法特点，从硬分类方法、软分类方法以及最新的软硬分类方法出发，总结了国内外的各研究成果，分析了各种方法的分类策略与特点及其方法适用性。研究结果表明：软硬分类方法能够灵活适用于遥感图像上纯净、混合像元并存的特点，可以有效解决光谱的异质性，在土地覆盖遥感监测中具有广阔的应用潜力。文中提出了基于变端元的软硬分类土地覆盖制图方法框架，并指出了今后的研究重点。

Review and prospect of land cover mapping by remote sensing

The study of land use/cover classification at regional scale is one of the important fields in the study of land use/cover change in the world. Accurate and timely acquisition of the characteristics of the earth's surface is essential to grasp the relationship and interaction between human and natural phenomena. Based on the characteristics of land cover remote sensing classification methods, this paper summarizes the research results here and abroad from hard classification methods, soft classification methods and the latest soft and hard classification methods, and analyses the classification strategies, characteristics and applicability of various methods. The results show that the soft and hard classification method can be flexibly applied to the characteristics of remote sensing images, such as the coexistence of pure and mixed pixels, and can effectively solve the spectral heterogeneity. It has broad application potential in remote sensing monitoring of land cover. In this paper, a framework of soft and hard classification land cover mapping method based on variable endpoints is proposed, and the research emphasis in the future is pointed out.