Date of Award
M.S. in Engineering Science
Computer and Information Science
This thesis presents a real-time Augmented Reality Navigation System(ARNavi) on Android smartphone that leverages the parallel computing power of mobile GPUs. Unlike conventional navigation systems, our proposed ARNavi augments navigation information onto real scene video streaming from device camera in real-time. To achieve this goal, we implement and accelerate compute intensive part of applications using OpenCL on GPU integrated on mobile Application Processor (AP). The contributions of this thesis are three-fold. First, we propose new lane detection algorithm and prediction mechanism based on geometric coordinates. The result shows that these two algorithms are fast and accurate. Second, we port and accelerate a complete application on mobile AP. By taking advantage of CPU-GPU heterogeneous computing techniques, we achieve more than 2.6 times performance boost compared to CPU only version. Lastly, we successfully integrate OpenCL and OpenCV on Android platform.
Zhao, Mengshen, "A GPU Powered Mobile AR Navigation System" (2015). Electronic Theses and Dissertations. 1298.
Emphasis: Computer Science