Displacement Measurements on Hyperelastic Materials Using Multiple Object Tracking: a Case Study

Soft robotics applications increasingly demand accurate measurement techniques for capturing detailed displacement fields in highly deformable, hyperelastic materials. Conventional methods, including strain gauge sensors, are limited in providing full-field strain data and may alter material stiffness, highlighting the need for alternative measurement approaches. This paper presents a methodology based on Multiple Object Tracking that employs paraboloid markers embedded in the surface of soft specimens to enable precise, contactless tracking of displacement fields. The methodology was tested on Dragon SkinTM specimens, mimicking artificial skin for robotic hands, and evaluated through compression experiments using an Instron Universal Testing Machine. Results show that the MOT algorithm yields an average standard deviation of $\sigma u=2.23 \times 10^{-2} ~\text{mm}$ in the horizontal direction and $\sigma v=1.78 \times 10^{-2} ~\text{mm}$ in the vertical direction, demonstrating promising accuracy relative to traditional techniques, with no significant impact on specimen stiffness from marker integration. This methodology overcomes the limitations of strain gauges by enabling reliable displacement measurements without altering specimen mechanical properties. It provides a valuable tool for validating finite element models, facilitating the refinement of constitutive models, mesh quality, and boundary conditions. Overall, the MOT-based methodology could enhance soft actuator characterization, providing highfidelity data essential for predicting mechanical responses in complex robotic applications.