Year: 2025
Type: conference-paper
Abstract: With the rapid advancement of satellite remote sensing technology, large-scale high-resolution raster imagery data has become a critical backbone for Geographic Information Systems (GIS), environmental monitoring, disaster early warning, and related fields. However, the continuous explosive growth of data volume and increasing demands for high real-time processing capabilities have highlighted significant performance and flexibility limitations in traditional storage solutions. To address these challenges, this study proposes an innovative distributed raster image processing framework optimized by memory caching, which integrates core modules such as cache management, memory virtualization, data persistence, and intelligent scheduling. By deeply integrating server memory resources with distributed computing architectures, the framework significantly enhances image data processing efficiency while achieving dynamic load balancing and multi-node collaborative optimization. Experimental results demonstrate that the proposed method substantially outperforms traditional disk-based solutions in terms of data throughput and task response speed, with an average improvement of 12% to 20% in processing efficiency. In addition, the framework design fully considers data security and persistence requirements, providing a systematic technical solution and practical foundation for high-real-time remote sensing image applications.
Authors: Xiao Yu, Xuehua Chen, Decheng Wang, Chunyan Song
Institutions: Chinese Academy of Sciences, Institute of Software
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4413845268