Human face tracking using Pose Estimation Algorithms in Augmented Reality

Augmented Reality is a new medium, combining aspects from ubiquitous computing, tangible computing, and social computing. This medium offers unique affordances, combining physical and virtual worlds, with continuous and implicit user control of the point of view and interactivity. Augmented Reality (AR) is a combination between computer-generated object and real-world object. Computer generated object is a result of a three-dimensional graphics. AR software is designed to provide real-time interactivity with the user. The implementation of AR in this paper is to developing spectacle frame model simulation with face tracking and pose estimation algorithm. This simulation will have a lot of benefits, since the user are able to try various models of spectacle frames, anytime, anywhere, and lead into effective technology implementation. The speed of face detection or face tracking used in this software will have a significant dependency from the resolution of image input. The lower the resolution, frame per second will be higher. After the software is activated, users that in the webcam’s range will be processed automatically. The spectacle frame model will be positioned in the user’s face. It has been found that the critical region-based processing steps could be parallelized, despite the resulting complex accumulation of intermediate results. The paper presents the parallel algorithms involved and the performance achieved. Comparison is made with more traditional edge-based systems, which may execute somewhat faster but are not as robust. The success of the parallelisation overcomes this performance limitation, and suggests a future production route.