The Framework of Cloud Computing Platform for Massive Remote Sensing Images

In recent years, due to the rapid development of remote sensing technology, a single high-quality image will occupy larger storage space, and video has become so widespread in the usage of environmental observation and record. Hence, digital data is growing exponentially, and how to manage them and make image processing more effectively is a key issue in Geographic Information System. Additionally, the limitation of hardware resource and time-consuming images' processing is a bottleneck to cope with such big data by commercial software in single PC. The aim of this paper is to propose a framework based on some standards of the interface (WCS, WMS, and WPS) from Open Geospatial Consortium (OGC), cloud storage from HDFS, and image processing from MapReduce. Within this framework, we implement image management as well as simple WebGIS and test a read/write performance under four kinds of data sets (Normal Distribution, Skew to Left, Skew to Right, and Peak in Left and Right). The results reveal write/read performance of HDFS are outperform than the local file system in the situation of larger files (most files range in size from 8 MB to 10 MB) and a large number of threads (threads equal to 40 or 50).