Decision-level fusion for single-view gait recognition with various carrying and clothing conditions
Image and Vision Computing, Volume 61, May 2017, Pages 54-69
Amer Al-Tayyan, Khaled Assaleh, Tamer Shanableh
Abstract：Gait recognition is one of the latest and attractive biometric techniques, due to its potential in identification of individuals at a distance, unobtrusively and even using low resolution images. In this paper we focus on single lateral view gait recognition with various carrying and clothing conditions. Such a system is needed in access control applications whereby a single view is imposed by the system setup. The gait data is firstly processed using three gait representation methods as the features sources; Accumulated Prediction Image (API) and two new gait representations namely; Accumulated Flow Image (AFI) and Edge-Masked Active Energy Image (EMAEI). Secondly, each of these methods is tested using three matching classification schemes; image projection with Linear Discriminant Functions (LDF), Multilinear Principal Component Analysis (MPCA) with K-Nearest Neighbor (KNN) classifier and the third method: MPCA plus Linear Discriminant Analysis (MPCA + LDA) with KNN classifier. Gait samples are fed into the MPCA and MPCALDA algorithms using a novel tensor-based form of the gait images. This arrangement results into nine recognition sub-systems. Decisions from the nine classifiers are fused using decision-level (majority voting) scheme. A comparison between unweighted and weighted voting schemes is also presented. The methods are evaluated on CASIA B Dataset using four different experimental setups, and on OU-ISIR Dataset B using two different setups. The experimental results show that the classification accuracy of the proposed methods is encouraging and outperforms several state-of-the-art gait recognition approaches reported in the literature.
The impact of speed and time on gait dynamics
Human Movement Science, Volume 54, August 2017, Pages 320-330
Kathleen S. Thomas, Daniel M. Russell, Bonnie L. Van Lunen, Sheri R. Colberg, Steven Morrison
Abstract：To determine the effects of speed on gait previous studies have examined young adults walking at different speeds; however, the small number of strides may have influenced the results. The aim of this study was to investigate the immediate and long-term impact of continuous slow walking on the mean, variability and structure of stride-to-stride measures. Fourteen young adults walked at a constant pace on a treadmill at three speeds (preferred walking speed (PWS), 90% and 80% PWS) for 30 min each. Spatiotemporal gait parameters were computed over six successive 5-min intervals. Walking slower significantly decreased stride length, while stride period and width increased. Additionally, stride period and width variability increased. Signal regularity of stride width increased and decreased in stride period. Persistence of stride period and width increased significantly at slower speeds. While several measures changed during 30 min of walking, only stride period variability and signal regularity revealed a significant speed and time interaction. Healthy young adults walking at slower than preferred speeds demonstrated greater persistence and signal regularity of stride period while spatiotemporal changes such as increased stride width and period variability arose. These results suggest that different control processes are involved in adapting to the slower speeds.
Development of a novel virtual reality gait intervention
Gait & Posture, Volume 52, February 2017, Pages 202-204
Anna E. Boone, Matthew H. Foreman, Jack R. Engsberg
Abstract：Introduction Improving gait speed and kinematics can be a time consuming and tiresome process. We hypothesize that incorporating virtual reality videogame play into variable improvement goals will improve levels of enjoyment and motivation and lead to improved gait performance. Purpose To develop a feasible, engaging, VR gait intervention for improving gait variables. Methods Completing this investigation involved four steps: 1) identify gait variables that could be manipulated to improve gait speed and kinematics using the Microsoft Kinect and free software, 2) identify free internet videogames that could successfully manipulate the chosen gait variables, 3) experimentally evaluate the ability of the videogames and software to manipulate the gait variables, and 4) evaluate the enjoyment and motivation from a small sample of persons without disability. Results The Kinect sensor was able to detect stride length, cadence, and joint angles. FAAST software was able to identify predetermined gait variable thresholds and use the thresholds to play free online videogames. Videogames that involved continuous pressing of a keyboard key were found to be most appropriate for manipulating the gait variables. Five participants without disability evaluated the effectiveness for modifying the gait variables and enjoyment and motivation during play. Participants were able to modify gait variables to permit successful videogame play. Motivation and enjoyment were high. Summary A clinically feasible and engaging virtual intervention for improving gait speed and kinematics has been developed and initially tested. It may provide an engaging avenue for achieving thousands of repetitions necessary for neural plastic changes and improved gait.