The Gender, Height and Weight Prediction Based on Gait Using MEMS Sensor

Human gait is one of the behavioral biological feature for personal identification and authentication. Gender, height, and weight information are also used as soft biological features in medical care, intelligent biological systems and other systems. However, the current research on the relationship between gait and the above three biological characteristics is mostly the analysis of single or two relationships. In this article, we propose a recognition and analysis method based on the fusion of gait data collected by a single wearable sensor with three human biological characteristics (gender, height, and weight). Perform complete gait segmentation based on pedestrian acceleration data and synchronized video, and extract important relevant gait feature values for intelligent prediction of three features by neural network. The results show that the method is easy to operate and low in cost. By establishing a correlation model between gait and the above characteristics, the accuracy of using gait to predict gender is 100%, and the effect of predicting height and weight is very good.