Abo Bibliothek: Guest
Digitales Portal Digitale Bibliothek eBooks Zeitschriften Referenzen und Berichte Forschungssammlungen
Critical Reviews™ in Biomedical Engineering
SJR: 0.26 SNIP: 0.375 CiteScore™: 1.4

ISSN Druckformat: 0278-940X
ISSN Online: 1943-619X

Volumen 48, 2020 Volumen 47, 2019 Volumen 46, 2018 Volumen 45, 2017 Volumen 44, 2016 Volumen 43, 2015 Volumen 42, 2014 Volumen 41, 2013 Volumen 40, 2012 Volumen 39, 2011 Volumen 38, 2010 Volumen 37, 2009 Volumen 36, 2008 Volumen 35, 2007 Volumen 34, 2006 Volumen 33, 2005 Volumen 32, 2004 Volumen 31, 2003 Volumen 30, 2002 Volumen 29, 2001 Volumen 28, 2000 Volumen 27, 1999 Volumen 26, 1998 Volumen 25, 1997 Volumen 24, 1996 Volumen 23, 1995

Critical Reviews™ in Biomedical Engineering

DOI: 10.1615/CritRevBiomedEng.2020035013
Forthcoming Article

Human Locomotion Classification for Different Terrains using Machine Learning

IIT (BHU) Varanasi
IIT (BHU) Varanasi
IIT (BHU) Varanasi
IIT (BHU) Varanasi


This paper presents gait analysis carried out on healthy subjects based on surface electromyographic and acceleration sensor signal, implemented through machine learning approaches. The surface EMG and 3-axes acceleration signals have been acquired for five different terrains- level ground, ramp ascent, ramp descent, stair ascent, and stair descent. These signals were acquired from the tibialis anterior and gastrocnemius medial head muscles that correspond to dorsiflexion and plantar flexion, respectively. After feature extraction, these signals are fed to five conventional classifiers- linear discriminant analysis, k-nearest neighbors, decision tree, random forest and support vector machine, that classify different terrains for human locomotion. We compared the classification results for above classifiers with deep neural network classifier. The objective of the present work is to obtain the features and classifiers that are able to discriminate between five locomotion terrains with maximum classification accuracy in minimum time by acquiring the signal from least number of leg muscles. The results indicated that support vector machine gives the highest classification accuracy of 99.20 (± 0.80) % for the dataset acquired from fifteen healthy subjects. In terms of both accuracy and computation time, the support vector machine outperforms other classifiers.