OpenCap: Movement Biomechanics from Smartphone Videos

Measures of human movement kinematics and dynamics can support rehabilitation decision making. However, these measures are rarely quantified in clinical practice due to the prohibitive cost, time, and expertise required. Here, we present and validate OpenCap, an open-source platform for computing kinematics and dynamics from smartphone videos. We collected data from 10 participants performing four types of activities (gait, squat, sit-to-stand, drop jump) and compared OpenCap estimates to a traditional marker-based motion capture approach. We found joint kinematic ( \(4.1^\circ\) ), ground reaction force (6.7% body weight), and joint moment (1.34% body weight times height) errors similar to those obtained with state-of-the-art solutions. In this study, we also show that OpenCap is accurate enough to inform rehabilitation decisions and efficient enough for quick in-field data collection from large cohorts. In the two years since its release, OpenCap has been used by over 5,000 researchers worldwide who collected over 150,000 motion trials.