Real-Time Cycling Cadence Estimation Using an Inertial Sensor for Gamified Pedaling Therapy

In an increasingly aging population, impaired motor control and loss of balance and lower limb muscle strength are triggering an increased risk of falls and hip fractures. With the intention of preventing them, are enforced intervention programs based on pedaling exercises to improve physical conditioning of the elderly. To achieve physical results adherence to cycling exercises must be promoted. For this reason, providing a compact and portable sensing solution to record pedaling parameters and offer visual real-time feedback would become very valuable. Therefore, obtaining pedaling information from the rotation measurements of an inertial sensor placed on the thigh, requires firstly the recognition of flexion/extension angle. Then, different processing steps are performed on the data consisting of a Kalman filter for smooth estimation of the angles, a peak detection consisting of the sliding window application and symmetry threshold of the angle curve. Finally, the performance of the cadence estimation method corroborates the suitability of its usage in cycling training feedback.