【摘要】 针对集中存储、分级存储存在的静态、粗粒度等问题，文中设计实现了遥感影像分层存储架构及其迁移方法。为了准确识别迁移对象，提出基于空间单元的遥感影像热度评估模型并顾及影像质量进行热度重排序，给出了分层存储迁移方法，实现了遥感影像数据的动态迁移。根据不同存储方式的性能差异和使用特点，构建了云存储与对象存储联动的分层存储架构。为验证方法有效性，模拟了遥感影像数据的访问和迁移过程，结果显示本方法相较于现有方法准确率提高9%以上，且稳定性更好，在高速存储空间有限的条件下能够提高存储利用率，实现遥感影像高效的存储管理。更多还原

【Abstract】 Aiming at the static and coarse-grained problems of centralized storage and hierarchical storage, a hierarchical storage architecture for remote sensing(RS) images and its migration method is designed and implemented in this paper. In order to accurately identify the images to be migrated, the RS image hotness evaluation model based on spatial units is proposed and the hotness reordering is carried out considering the image quality, and the hierarchical storage migration method is given to realize the dynamic migration of RS image data. According to the performance differences and usage characteristics of different storage methods, a hierarchical storage architecture with the linkage between cloud storage and object storage is built. To verify this method, the access and migration process of RS image data is simulated. The results show that the accuracy of this method is 9% higher than the existing methods, and the stability is better. It can improve the storage utilization rate and realize the efficient storage management of RS images under the condition of limited high-speed storage space.更多还原