A Novel EMG-Based Stride Length Estimation Method for Pedestrian Dead Reckoning

W. Chen, X. Zhang

Abstract:	Generally, to bridge the gaps when GPS outages occur, traditional Pedestrian Dead Reckoning (PDR) solutions measure acceleration from accelerometers to derive the speed, and the Earth’s magnetic vector from digital compasses or angular rate from gyros to obtain the walking heading. However, these signals are sensitive to instrumental errors and disturbances from ambient environments. To be totally different from these signals in nature, electromyography (EMG) signal is a typical kind of biomedical signal which measures electrical potentials generated by muscle contractions from human body, not the physical quantity with respect to the environment. Since this kind of signal would reflect muscle activities during human locomotion, it can not only be utilized for speed assessment, but also disclose the azimuth information from the contractions of lumbar muscles during postural adjustments or rhythmic activities. Considering the advantages that the EMG signal can be used for obtaining two different kinds of information (speed and azimuth) without being sensitive to environments, and that the potential applications would be in comprehensive context awareness (for example, getting the position and gesture altogether), we introduce this signal firstly into pedestrian navigation. The work presented in this paper is our first step in utilizing the EMG signal for stride detection and stride length estimation in PDR. Since during walking, the contractions of muscles are cyclic and the intensity of the EMG signal directly represents the force exerted by the legs, the stride occurrence can be detected and the corresponding stride length is estimated via a linear model of stride period and stride peak (the maximum value of EMG signal during current stride). To validate the feasibility and effectiveness of this method suggested, several field tests were conducted by three testers. The results demonstrate that the proposed method can detect more than 99% of stride occurrences and the average error of the estimated stride length is lower than 5 cm. In addition, three simulated GPS gaps were intentionally introduced to evaluate the performance of this method in PDR, with the heading from a 2-axis digital compass. The results indicate that the performance of the EMG-based PDR is comparable to that of GPS under open-sky environments.
Published in:	Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010) September 21 - 24, 2010 Oregon Convention Center, Portland, Oregon Portland, OR
Pages:	569 - 576
Cite this article:	Chen, W., Zhang, X., "A Novel EMG-Based Stride Length Estimation Method for Pedestrian Dead Reckoning," Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010), Portland, OR, September 2010, pp. 569-576.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In