Stride Length Estimation using Fuzzy Logic and a Single Shoe Mounted IMU Sensor

Stride length estimation is a problem in pedestrian dead reckoning. Accurate stride length is important since it yields precise pedestrian localization. This paper presents a new method to estimate stride length by using Fuzzy logic and a single IMU sensor mounted on shoe as well as a method to detect a stride. The IMU sensor measures foot acceleration signal and the different foot acceleration between the measured value and a collected value is used as input signal for the proposed Fuzzy logic design. Based on a 100 meters long walk, the subjects are divided into three groups by their walking style whereas the Fuzzy logic rules are designed for each type. The proposed method is different from existing method in that it needs only the foot acceleration measured data and does not require subject height or stride frequency. The experimental results of 11 subj ects show that the stride detection and also the Fuzzy logic are successfully implemented.