顾及时空语义的多主题瓦片数据优化检索方法

针对虚拟地球可视化中多个逻辑图层叠加产生大量同名瓦片索引重叠引起的数据无效访问和内存冗余等问题,提出顾及时空语义的瓦片数据优化检索方法。在客户端实现顾及时空语义的自适应瓦片优选,将视点信息与数据集的时空范围、分辨率和优先级等语义信息自动匹配与自适应筛选过滤,显著提高目标瓦片数据的命中率;在服务器端实现面向主题的内存数据库瓦片数据缓存,进一步提高瓦片数据二次访问的响应速度。实验表明,该方法不受逻辑图层增量影响,保证目标瓦片较高的命中率,缓存方法进一步提升瓦片二次访问效率,显著提高面向多数据集的海量数据实时可视化性能。

An optimal retrieval method of multi-theme image tiles considering the spatio-temporal semantics

Aiming at the invalid accesses and memory redundancies caused by data overlap of more than one logic map tile in the visualization process of virtual earth, an optimal retrieval method of multi-theme tiles considering the spatio-temporal semantics is proposed. At the client side, the self-adoptive retrieval of tiles considering spatio-temporal semantics automatically matches the viewpoint information with semantics of datasets like space-time range, resolution and priority, then it filters the invalid selection and improves the efficiency of target tiles access significantly. At the server side, the theme oriented data cache based on memory database is achieved to prompt the response speed of server in the second access of tiles. At last, the experiments prove that this method could prevent stable and high hit-rate of target tile affected by amount of datasets and the cache method further improve the efficiency of tiles access. The real-time visualization performance of massive image data is significantly improved.