An fast integrated searching strategy and application in multi-source massive image database for Disaster Mitigation and Relief

Disaster Mitigation and Relief (DMR) has been paid much more attention from national or local governments in China. Multi-source remotely sensed data with different spatial, spectral and temporal resolutions were widely applied in DMR (e.g. Disaster Warning, Rapid Response, Loss Evaluation and Reconstruction, etc.), and have been a main data source in the database of DMR Although some famous remote sensing image database engines have the power to organize, store and manage massive image data, their searching efficiency in massive imagery is poor because of the fact that the temporal, spatial and spectral attributes were not taken account in their image database construction procedure. Especially for the preprocessed imagery used in DMR, the common image management software packages usually store and manage them as individual scene depending upon the catalog. It leads to a low efficiency for searching and accessing images. In this study, a novel fast integrated searching strategy (FISS) was proposed and a set of grid indexes were designed to project multi-dimensional image data with multi-temporal, multi-spectral and multi-spatial resolutions to the 2-D space. The required image data at different locations for any acquired dates, spectral and spatial resolutions could be rapidly searched by different dimensional indexes. In order to improve the speed and efficiency of data access, the preprocessed images were divided into tiles and compressed. FISS makes the system be interactive and more convenient to the users. It has been used in disaster mitigation and relief database of disaster reduction in China. In this paper, the proposed searching technique was compared with traditional ones, the results showed that FISS could accelerate searching speed, shorten disaster emergency response time.