Real-time Visualization of Vector Data Projected on Large-Scale Virtual Terrain Environments for Web-based GIS Frameworks

One of the key components of geographic information systems (GIS) is vector geospatial data visualization. Client-side vector data visualization in web-based GIS can be performed with geometry-based methods, projection-based methods or decals. Geometry-based methods and decals deliver high quality images, but display a decrease in performance on large amounts of dynamic vector data due to WebGL graphic web-API limitations. Projection- based methods present a suitable alternative with high performance but lower quality due to aliasing. Many projection-based algorithms focusing on reducing aliasing have been developed in shadow rendering. In this study, we aim to adapt projection-based shadow rendering algorithms for client-side vector data visualization in web-based GIS. As the result of our work, we present several such adaptations of projection-based methods, an analysis of their performance and precision, and a comparison of projection-based methods and decals. The practical implication of our work is the conclusion that projection-based vector data visualization methods are more suited for visualization of large amounts of dynamic vector data in web applications due to higher performance without the loss of image quality. Keywords: real-time rendering, vector data visualization, projective transformations, decals, WebGL, three.js.