We present a motion based adaptive step length estimation algorithm for PDR system using only a smartphone. It is difficult to estimate step length of user using only a smartphone because various motions are existed on smartphone usage environment and smartphone is widely used. To solve problem, we apply a motion recognition algorithm and gait cognition algorithm on step length model. Map based gait cognition algorithm enhances the performance of motion based step length estimation algorithm. To demonstrate the performance of proposed step length estimation algorithm, we conduct a field test in indoor and outdoor. A performance of basic process of PDR present that proposed algorithm can be effectively operated on smartphone based PDR system.

Keywords: motion, smartphone, PDR, step length