Digital image correlation (DIC) is a powerful, non-contact technique for quantifying strain fields, yet its accuracy in materials undergoing large deformations remains a challenge due to decorrelation, feature loss, and strain definition sensitivity. This study investigates the application of incremental DIC to capture the thermomechanical response of an open-cell polyurethane foam subjected to compressive strain 65% at temperatures ranging from 10 \(^{\circ }\) C to 40 \(^{\circ }\) C and strain rates between 10 \(^{-3}\) and 10 \(^{-1}\) s \(^{-1}\) . Conventional and incremental DIC approaches were compared using multiple strain definitions, revealing that global extensometer-based measurements underestimated strain heterogeneity by up to 15%. The Biot strain provided the most reliable bulk deformation representation, while a localized small strain definition improved the spatial resolution of strain gradients. Incremental DIC mitigated decorrelation effects at large strains, maintaining higher correlation coefficients, but introduced accumulated image noise over successive frames. These findings underscore the need for careful selection of the definition of strain and experimental design when applying DIC to highly deformable, energy-absorbing materials, informing improved methodologies for protective system development.
