Gait Symmetry Analysis Based on Affinity Propagation Clustering

When we are conducting an investigation in gait symmetry analysis, we usually grouped the test objects by decade intervals. However, this artificial division method has a large defect that does not truly reflect the human relationship between age and physical condition. Therefore, we found a new grouping method while clustering on the values of the difference in gait symmetry. The reason for using the affinity propagation clustering algorithm was that it can do clustering on data while passing the original information not the random values to processing clustering. And meanwhile this algorithm has good performance and efficiency. It will help us to do gait analysis more effectively, and also more reasonable to explain the relationship of human gait and age, or gait and other characteristics.