基于自适应反射率融合模型的遥感影像分辨率提高方法研究

Research on remote sensing image resolution improvement method based on adaptive reflectivity fusion model

遥感具有观测范围广,时效强的特点,广泛应用于铁路工程之类的大面积高复杂地区的监测。然而由于技术和成本的制约,难以获取长时序高分辨率的影像,制约了遥感对地监测的发展。本文基于时空自适应反射率融合模型(ESTARFM),在Google Earth Engine(GEE)云平台对其进行实现。分别选取了黑龙江省份的农村、城市、城郊结合类型的三个区域进行了实验,验证了该算法在预测大面积多变地物的反射率的准确性。结合该算法和平台,能以较快的速度得到高时空分辨率的连续的时间序列影像,满足遥感快速大规模动态监测陆地表面信息的需要。

Remote sensing has the characteristics of wide observation range and strong timeliness,and is widely used in monitoring large-scale and highly complex areas such as railway engineering.However,due to technical and cost constraints,it is difficult to obtain long-term high-resolution images,which restricts the development of remote sensing earth monitoring.This paper is based on the spatiotemporal adaptive reflectivity fusion model(ESTARFM)and is implemented on the Google Earth Engine(GEE)cloud platform.Three areas of rural,urban,and suburban areas in Heilongjiang Province were selected for experiments,which verified the accuracy of the algorithm in predicting the reflectivity of large-area changing ground objects.Combining this algorithm and platform,continuous time series images with high spatial and temporal resolution can be obtained at a faster speed,meeting the needs of remote sensing for rapid large-scale dynamic monitoring of land surface information.