Research On Acquisition And Retrieval Method Of The Locatable Image

As a kind of multimedia data, image was the reflection of the real world, by which the described geographic space strongly presents a sense of reality to provide an easier way gaining information. Nonetheless, it was cannot to pick up the position information via image, hence it cannot be used as a spatial data. Since the concept of the locatable image had been proposed, the locatable image, as a brand new expression of geographic space, can be defined as an integration of both image and space data that obtains an image with space information, and it had been the focus of GIS research.In this paper, a new data expression model of the locatable image was proposed. As the successor of locatable image, it increase many multi-source heterogeneous data, such as azimuth angle, attribute information and audio, to expand the geographical expression; Analyzed the advantaged and architecture of the locatable image model in detail, and studied the acquisition and fusion process of the image, geographical position, azimuth, user-defined worksheet and audio, thus to empirically verify the effectiveness of that model; Introduced a concept of visibility analysis, using the parameters in the locatable image model, such as geographical position, azimuth and focal length, to build a visual field model of image; The high efficiency of spatial index had been verified by experiments with using the multistage grid index of SQL Server to indexing geographical location and visual field of the locatable image. Furthermore, following conclusions had been drawn:the overall efficiency of index based on 8*8 grad was high, and the retrieved result based on visual field was more likely to meet the retrieving purpose; At last, a customizable prototype system for field data acquisition and retrieval had been developed based on Android and ArcObjects platform.