Research on Human Gait Characteristics at Different Walking Speeds Based on Alphapose

摘要

The gait characteristics of normal people have important reference significance for the research and development of lower limb exoskeleton rehabilitation robot. In order to solve the problem that the devices in traditional gait feature research is difficult to popularize, a human gait feature research method based on human posture estimation algorithm AlphaPose at different walking speeds is proposed. Firstly, the human skeletal joint point coordinates at different walking speeds were extracted using AlphaPose, a human pose estimation algorithm; Then, the human knee joint angle in a continuous frame picture is calculated according to the cosine theorem; Next, the maximum flexion angle in the loading response phase, the minimum flexion angle in the stance phase, and the maximum flexion angle in the swing phase of all subjects at the same walking speed were averaged respectively; Finally, the human gait characteristics were analyzed by the effect of walking speeds on various angles. The experimental results showed that the maximum flexion angle in the loading response phase, the minimum flexion angle in the stance phase, and the maximum flexion angle in the swing phase were significantly different (P < 0.05) at walking speeds of 2 km / h, 4 km / h, and 6 km / h; With the increase of walking speed, various angle values increase. This method is simple and effective, and can be used in the study of human gait characteristics at different walking speeds, to provide a reference basis for the design of lower limb exoskeleton rehabilitation robots.