Real Time Three Dimensional Motion Tracking Using Smartphones
Software that requires users to explore virtual three dimensional spaces, such as modeling/CAD programs and videogames, is becoming increasingly common. Despite the prevalence of such software, interfaces have remained stagnant and failed to evolve. Many still attempt to map two dimensional inputs to a three dimensional output which often results in user confusion and an increased learning curve. We propose a new type of interface that utilizes natural gestures to provide a more intuitive way to explore 3D spaces. Using smartphones, we have created a system that is able to track the devices' movement in three dimensions, without the need for additional sensors or prior setup. It takes advantage of the various sensors found in modern phones as well as the camera to estimate motion. We detail the motion estimation pipeline, including its strengths and weaknesses.