An SVM-Based Pedestrian Gait Recognition Algorithm Using a Foot-Mounted IMU

Based on the real-time data of angular velocity and acceleration sensed by IMU, this paper uses the machine learning method, support vector machine (SVM), to find a way to recognize and classify some common action types such as walking, running, going upstairs, going downstairs, jumping, etc. after data preprocessing and dimensionality reduction with Principal Component Analysis (PCA). The algorithm provides a basis for identifying zero speed at different gait states to improve the relevance of the zero-speed correction algorithm. And it will help identify different pedestrian motion states as a basis for motion constraint algorithms.