Gait segmentation using bipedal foot pressure patterns

摘要: We present an automated gait segmentation method based on the analysis of foot plantar pressure patterns elaborated from two wireless pressure-sensitive insoles. The 64 signals recorded by each device are to extract 10 feature variables which used segment cycle into 6 sub-phases following a simplified version Perry's model. is Hidden Markov Model with minimum phase length constraint and univariate Gaussian emission model, decoded using classic Viterbi algorithm. tested pool 5 healthy young subjects walking at different speeds, through leave-one-out cross-subject validation. results show that highly effective, yielding average performance about 95% correct classification, 85 90% transitions detected inside acceptance window 50ms.