• Journal of Institute of Control, Robotics and Systems
• /
• v.18 no.7
• /
• pp.644-649
• /
• 2012

This paper proposes the gait phase classifier using backpropagation neural networks method which uses the angle of lower body's joints and ground reaction force as input signals. The classification of a gait phase is useful to understand the gait characteristics of pathologic gait and to control the gait rehabilitation systems. The classifier categorizes a gait cycle as 7 phases which are commonly used to classify the sub-phases of the gait in the literature. We verify the efficiency of the proposed method through experiments.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2013.05a
• /
• pp.321-322
• /
• 2013

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.6
• /
• pp.518-523
• /
• 2015

This paper proposes a gait phase classifier using a Recurrent Neural Network (RNN). Walking is a type of dynamic system, and as such it seems that the classifier made by using a general feed forward neural network structure is not appropriate. It is known that an RNN is suitable to model a dynamic system. Because the proposed RNN is simple, we use a back propagation algorithm to train the weights of the network. The input data of the RNN is the lower body's joint angles and angular velocities which are acquired by using the lower limb exoskeleton robot, ROBIN-H1. The classifier categorizes a gait cycle as two phases, swing and stance. In the experiment for performance verification, we compared the proposed method and general feed forward neural network based method and showed that the proposed method is superior.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.2
• /
• pp.17-26
• /
• 2018

In this paper, we propose a gait type classification method based on pressure sensor which reflects various terrain and velocity variations. In order to obtain stable gait classification performance, we divide the whole gait data into several steps by detecting the swing phase, and normalize each step. Then, we extract robust features for both topographic variation and speed variation by using the Null-LDA(Null-Space Linear Discriminant Analysis) method. The experimental results show that the proposed method gives a good performance of gait type classification even though there is a change in the gait velocity and the terrain.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.3
• /
• pp.181-189
• /
• 2015

Powered prosthesis is used to assist walking of people with an amputated lower limb and/or weak leg strength. The accurate gait phase classification is indispensable in smooth movement control of the powered prosthesis. In previous gait phase classification using physical sensors, there is limitation that powered prosthesis should be simulated as same as the speed of training process. Therefore, we propose EMG signal based gait phase recognition method to classify stairs ascending and stairs descending into four steps without using physical sensors, respectively. RMS, VAR, MAV, SSC, ZC, WAMP features are extracted from EMG signal data and LDA(Linear Discriminant Analysis) classifier is used. In the training process, the AHRS sensor produces various ranges of walking steps according to the change of knee angles. The experimental results show that the average accuracies of the proposed method are about 85.6% in stairs ascending and 69.5% in stairs descending whereas those of preliminary studies are about 58.5% in stairs ascending and 35.3% in stairs descending. In addition, we can analyze the average recognition ratio of each gait step with respect to the individual muscle.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.1
• /
• pp.6-11
• /
• 2011

In this paper, we propose a gait-based human identification system using eigenfeature regularization and extraction (ERE). First, a gait feature for human identification which is called gait energy image (GEI) is generated from walking sequences acquired from a camera sensor. In training phase, regularized transformation matrix is obtained by applying ERE to the gallery GEI dataset, and the gallery GEI dataset is projected onto the eigenspace to obtain galley features. In testing phase, the probe GEI dataset is projected onto the eigenspace created in training phase and determine the identity by using a nearest neighbor classifier. Experiments are carried out on the CASIA gait dataset A to evaluate the performance of the proposed system. Experimental results show that the proposed system is better than previous works in terms of correct classification rate.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39B no.7
• /
• pp.479-490
• /
• 2014

A lower extremity exoskeleton is a robot device that attaches to the lower limbs of the human body to augment or assist with the walking ability of the wearer. In order to improve the wearer's walking ability, the robot senses the wearer's walking locomotion and classifies it into a gait-phase state, after which it drives the appropriate robot motions for each state using its actuators. This paper presents a method by which the robot senses the wearer's locomotion along with a novel classification algorithm which classifies the sensed data as a gait-phase state. The robot determines its control mode using this gait-phase information. If erroneous information is delivered, the robot will fail to improve the walking ability or will bring some discomfort to the wearer. Therefore, it is necessary for the algorithm constantly to classify the correct gait-phase information. However, our device for sensing a human's locomotion has very sensitive characteristics sufficient for it to detect small movements. With only simple logic like a threshold-based classification, it is difficult to deliver the correct information continually. In order to overcome this and provide correct information in a timely manner, a probabilistic gait-phase classification algorithm is proposed. Experimental results demonstrate that the proposed algorithm offers excellent accuracy.