基于遥感的冰川信息提取方法研究进展

Study of Extracting Glacier Information from Remote Sensing

对冰川监测中常用的遥感卫星、传感器及冰川信息提取方法等进行了综合评价,常规方法中普遍认为比值法的精度最高,新产生的面向对象分类和雷达干涉测量方法虽一定程度上提高了冰川提取精度,但冰碛物仍是自动识别的难点.针对表碛覆盖冰川虽发展了一些自动、半自动的方法,但这些方法还不够成熟、不具有通用性.积雪、冰碛物和地面验证仍是冰川自动提取存在的重要问题,发展更先进、更成熟的方法是冰川研究的重要方向,未来可以尝试采用粗糙集理论及ICESAT卫星波形提高冰川信息提取的精度.

In this paper,remote sensing satellites,sensors used in monitoring glacier change and methods of extracting glacier information are summarized.Various approaches are evaluated synthetically.It is considered that the method of band ratio is the best among the conventional methods.The new methods of object-oriented classification and SAR interferometry have improved the accuracy of extracting glacier information to a certain extent.However,superglacial debris is still a difficulty in automatic recognition of glacier information.Automatic and semi-automatic methods have been developed for debris-covered glaciers.However,the methods are immature and out of universality.Obstacles of snow cover and superglacial debris and hard ground verification of glacier information derived from satellite images still need to overcome.Therefore,to develop more advanced and more sophisticated methods will be an important issue in glacier research.It is expected that rough set theory and waveform of ICESAT satellite will improve the accuracy of extracting glacier information.