Theory And Technology Research Of 6DOF Self-Tracking Using Smartphones

Self-tracking with smartphones at 6 degrees of freedom in real-time is essential in lots of applications and other fields such as Virtual Reality.With the development of built-in sensors in smartphones,many self-tracking solutions have appeared.Recently,many researchers try to utilize vision-based approaches(SLAM,Simultaneous Localization And Mapping)combined with Inertial Measure Unit(IMU)to realize self-tracking with smartphones.After testing these approaches,however,we find that the tracking can be lost in the following four scenarios:(1)rotating IMU fast for a long period of time will cause serious delay of orientation tracking;(2)the scene where background features are not distinct enough;(3)when smartphone moves fast,image features become very different in successive frames;(4)unstructured scene where background features are not static.To address these issues,we propose self-Tracker,which utilizes Real-time StepLength Adaption Algorithm aided by Gyroscope-compensation method to solve scenario(1)and Parallel-Multi-State Local Recovery method to deal with scenarios(2)-(4).Extensive experiments show that self-Tracker realizes robust and accurate self-tracking in these four scenarios with an error of 0.7% during the whole trajectory.Meanwhile,we design a gesture recognition method based on self-Tracker system which can achieve the same or higher accuracy than original recognition method with less samples.Through experiments,we find that our method has over 70% accuracy when the sample number of one gesture is just 2.