Fast visualisation of massive data based on viewpoint motion model
Chang Bao Wen, Lu Li, Kai Xie · Electronics Letters · 2017
To display massive data accurately and in real time, a visualisation algorithm based on the viewpoint motion model is proposed. A Hilbert‐R tree is constructed to index massive data. Further, the massive data is divided into a visible area, a potential area, and an unloading area using the viewpoint motion model. According to the index, corresponding data blocks are loaded into the memory, and data blocks of the visible area further transferred into the graphic processing unit (GPU) memory for rendering. When the viewpoint is changed, the data block of the potential area in the memory is directly loaded into the GPU. Experimental results show that under the premise of ensuring the quality of rendering, massive data can be displayed interactively, and the speed of displaying increases by 36–78%.