Volume visualization from compressed data in an Android device
Tutor / director / evaluatorVázquez Alcocer, Pere Pau
Document typeMaster thesis
Rights accessRestricted access - confidentiality agreement
GPU-Volume rendering had a great impact in the way of visualizing multidimensional data because of the great parallelism that this architecture provides. Since the programmable GPUs appeared various techniques of volume rendering have been implemented to exploit the parallel processing capabilities of this architecture. Rendering volume models are still a challenge and this is caused by two main problems: first, the intensity of computation that requires the typical algorithms currently implemented in GPUs and second, the fact that by its nature, volumetric data is large and often beyond the amount of information that can be stored in real-time, either in GPU and even RAM memory. This is why many researchers have developed techniques to mitigate these issues. Recently mobile devices have considerably increased their processing power. Scientists and developers have taken interest in creating scientific visualization in these platforms, but creating applications in these platforms is quite different compared to the traditional applications that are developed for desktops, workstations or laptops, because mobile devices have limited memory and processing power. Another factor to take in account is the smaller screen that these devices have, this leads to the necessity to adapt or create graphical interfaces in an efficient manner. This thesis objective is to develop an application that is able to render volumetric datasets at interactive frame-rates and decrease the memory footprint in mobile devices using a variety of texture compressions that mobile devices support nowadays.
|Federico Valencia.pdf||2.719Mb||Restricted access|