| The geographic information expression is not only a fundamental problem in geography, but the core content in the field of virtual geographic environment research as well. But so far, the visualization of VGE still faces challenges in many fields like contradictions of embodiment and abstract, popularity and specialization, thus practical needs cannot be satisfied by only relying on traditional vector and raster data, three-dimensional geometrical modeling, and professional way of obtaining data. Therefore, it's especially important for VGE construction of exploring a new visualization method and data acquisition method to meet the characteristics of human cognition and visual habits. The three-dimensional panorama technology has significant scientific meaning and application value to the construction of VGE. As it is advanced in complexity of model irrelevant to the complexity of the scene, less calculation, lower dependence on hardware, high immersion and authenticity, 360 ° panoramic expression of the real world, thus it has potential application value to VGE, it forces VEG rise to the geographic space expressing level based on panoramic data by breaking through traditional expression methods based on vector and raster data and three-dimensional geometry models.This paper focused on time-space modeling of VGE based on panoramic data, and mainly researched on panoramic data acquisition, geographic panorama data model design and the time-space model of VGE based on panoramic data, and then a prototype system was designed to validate relevant theories and methods in this paper. The main works of this research were as follows:(1) Introduced concepts of “Crowdsourcing Geographic Panoramic Data ” and “Crowdsourcing Panoramic Virtual Environment”, and designed a new fast panoramic stitching algorithm based on mobile devices. Relevant concepts were defined on the basis of GIS socialization and VGI which inherited and developed from the concept of “user spontaneously data gathering, managing and applying” of socialization and VGI. Public geographic spatial data and crowdsourcing geographic panoramic data into VGE were introduced, public data gotten windows were opened, that is, the acquisition, processing, storage, sharing of the panoramic data can break through traditional data acquisition mode and delivered it to users who have double identity of users and data contributors with the data sharing platform to supplement data source of VGE, expected to escape single data source and turn into popular and universal enforceable. The features and key technologies of the crowdsourcing geographic panoramic data were analyzed, and the relation and distinction of the crowdsourcing geographic panoramic data and spontaneous geographical information were elaborated from the aspects of concepts, information carrier and way of public participation.To create panoramic VGE, especially under situation of relatively larger space scale, the support of huge amount of panoramic data was needed which not only collected by professional equipment but public collaborative data collection as well. Generally, the panoramic data can be collected by general digital camera, CCD camera which growing into mature and panorama mosaic algorithm used on normal computer as well. With the development of mobile phone and tablet computer in hardware parameters in recent years, many types of equipment have functions of generating panorama pictures, even so, they can't match normal computers either on hardware parameters or calculations, in addition, data collectors' professional level is uneven, therefore, quite many panoramas quality are not satisfied, and the special mobile device panorama generation algorithms have been less researched, the algorithms of normal computer can't be simply used on mobile devices but according to features of itself and make out its own algorithms.This paper designed a new faster panoramic stitching algorithm. Compared to traditional methods, the new method used color correction to reduced color differences which can shorten the time of calculating equation in the late period of image fusion. Additionally, the staying stitching images and composite images were all loaded into memory in traditional methods, whereas in new algorithm, low resolution images loaded into memory while high resolution images loaded into hardware blocks, in this way, the new method takes less memory. Eventually, factors of moving objects were considered in this algorithm, which can remove ghosts relatively thoroughly compared to traditio nal algorithms.(2) Reseached on geographic panoramic data model, data index, geographic panoramic data processing method. From perspective of integration, standardization and scalability, practicability and versatility, a designing principle for geographic panoramic data model was made on the basis of panorama data model, spatial geometric data model as well as semantic annotation data model design, then the model diagrams were drawn, semantic relationships between objects and components were analyzed, the model functions and design goals discussed. The data retrieval method upon location and spatial relations and semantics was explained and analyzed, finally, external data sources, data acquisition and processing, means of geographic panoramic database creation were discussed.(3) Researched on time-space model of crowdsourcing panoramic virtual environment. The ultimate aim of crowdsourcing panoramic virtual environment was making users perceive geographical phenomenon, performing dynamic simulation analysis, monitoring the change process of time and space, understanding and transforming geographic world in collaborative interaction way in panoramic space, to achieve these goals, the time-space model of crowdsourcing panoramic virtual environment should be researched. We defined concepts of “Geographic Panoramic Unit”, “Panoramic Geographic Space”, “Geographic Panoramic Unit Coupling”, and researched basic structure, behavior modeling characteristics, relationship of time and space and space information of geographic panoramic space, and method of creating crowdsourcing geographic panoramic data based on VGE.The thesis analyzed the relationship between consistency of panoramic geospace and geographic panoramic unit, gave image array spatial interpolation coupling method and the image linear integral convolution coupling method. As the basic element of panoramic geographic space, the coupling degree of geographic panoramic unit is a significant discriminant mark of panoramic geographical spatial consistency, and affects the quality of construction of panoramic VGE. Therefore, the thesis developed different coupling methods for different space scales. Image array interpolation method was presented for smaller space scales. The image linear integral convolution coupling method was introduced as to larger ones, by this means, a kind of time based panoramic geographic space coupling method was realized.(4) Prototype system verification. This thesis took Beijing for example and collected and processed geographic panoramic data, developed panoramic VGE experimental platform of Beijing and public panorama data sharing platform on mobile device, verified the validity and effectiveness of the related theories in this paper. |
PDF Full Text Request