A novel approach for discrimination of human gait using kernel learning algorithm

The discrimination of human gait change has important role in clinical and rehabilitation contexts. In order to avoid limitations of the quantitative assessment of the change of human gait such as the loss of the useful gait feature information and the lower sensitivity, kernel-based linear discriminant analysis (LDA) for recognizing change of human gait was investigated. The gait data of two groups of participants including young and old peoples respectively were acquired during normal walking, and were analyzed using the proposed method. Experimental results indicated that the proposed technique could obtain more useful gait feature information related to human gait than LDA, and identify gait patterns with 87.5% accuracy, demonstrating that the proposed technique could be used as an effective tool for the discrimination of human gait in the future clinical context.