Cloud Computing Based on Computational Characteristics for Disaster Monitoring

Applied Sciences

卷:10

Wenyang Yu¹  Guoqing Li¹  Quan Zou² 

[1] Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China;

[2] School of Computer Information and science, Centre for Research and Innovation in Software Engineering, Southwest University, Chongqing 400715, China;

关键词: programming model;    disaster monitoring;    computational characteristics;    remote-sensing data;   

DOI  :  10.3390/app10196676

来源: DOAJ

【 摘 要 】

Resources related to remote-sensing data, computing, and models are scattered globally. The use of remote-sensing images for disaster-monitoring applications is data-intensive and involves complex algorithms. These characteristics make the timely and rapid processing of disaster-monitoring applications challenging and inefficient. Cloud computing provides a dynamically scalable resource over the Internet. The rapid development of cloud computing has led to an increase in the computational performance of data-intensive computing, providing powerful throughput by distributing computation across many distributed computers. However, the use of current cloud computing models in scientific applications using remote-sensing image data has been limited to a single image-processing algorithm rather than a well-established model and method. This poses problems for the development of complex disaster-monitoring applications on cloud platform architectures. For example, distributed computing strategies and remote-sensing image-processing algorithms are highly coupled and not reusable. The aims of this paper are to identify computational characteristics of various disaster-monitoring algorithms and classify them according to different computational characteristics; explore a reusable processing model based on the MapReduce programming model for disaster-monitoring applications; and then establish a programming model for each type of algorithm. This approach provides a simpler programming method for programmers to implement disaster-monitoring applications. Finally, some examples are given to explain the proposed method and test its performance.

【 授权许可】

Unknown