Geo-tagged image based on Voronoi and TIN module

Abstract

The current increasing use of Geo-tagged data is creating new demand for web mapping service applications. Until now, most mapping service web pages have already used Geo-tagged elements, such as Google Map, Microsoft’s Bing Map and Nokia’s Here Map. It has also been noticed that the number of researchers interested in Geo-tagged Map has increased recent years. Some are concerned with establishing local lexicons for retrieving data accurately, while others want to know more about how to retrieval geo-tagged images efficiently. There are also many researchers interested in how to represent geo-tagged images. However, the challenge of representing images is that since all the online data sets are unstructured, it is necessary to find sophisticated algorithms to represent large image data sets as graphs or networks, where each image is a vertex and is connected to surrounding other images, thus the whole dataset can be easily deal with. In this paper, two most popular algorithms, triangular irregular network and voronoi structure are discussed and applied as basic structures to display images (there is no such research until now) in both 2D plan maps and 3D sphere from Google Earth. It is no double that there are significant differences between these two platforms. More specifically, 2D structure has highly developed, while 3D structures are less mature. In this paper, the limitations and potential improvements to these algorithms will be described, and future work will be demonstrated. Results from the evaluation questionnaire indicate that the program interface and colour choice are favoured by most volunteers, while the efficiency and cross-browser compatibility have major problems and need to be refined in the future. The assessment of preference of effects demonstrated that the “lomo” approval is inferior to the other five results for producing a 2D plan map.