Abstract
Walking patterns can be used as a key parameter in identifying individuals, as it varies visually depending on one's body size as well as their habits, gender, and age group. In this study, we measure the gait characteristics of a large number of subjects using 34 visual parameters to identify significant parameters that can be used to distinguish individual walking features. We recorded 291 subjects' walking on a constructed footpath using four video cameras, and data on parameters was calculated at the points of double support, toe-off, and heel-strike. K-means Clustering Analysis and ANOVA were conducted to determine the difference between age, gender, and BMI. As a result, we confirm that parameters related to the spine, neck, and feet are useful for identifying individuals. In the comparative analysis between age groups, the older the age, the more significant variables appeared in the upper body. The difference between genders showed significant parameters in both the upper and lower bodies of males. Similarly, among the large BMI groups, we also derived significant results in the upper and lower bodies. The key parameters derived from this study can be used more effectively in the real-world visual analysis of gait, as the walking characteristics of a large number of subjects have been measured with a similar view as real-world CCTV. This study will be effectively utilized as a foundation for future research attempting to identify people through their gait by distinguishing major gait characteristic differences.