Abstract
In the user-centered design of hand-product interfaces, knowledge of the hand’s dimensions is of importance. While extensive databases and digital models exist, accurate measurements of individual users’ hands are required in certain cases. Combining an AI-based landmark recognition for hands with contour recognition and a hardware setup, this work introduces a measurement method for two-dimensional hand surface measurements. An experiment with n = 30 participants is performed to compare this method to the established standard (manual caliper) in anthropometric measurements following Bland-Altman. As a result, the image-based measurement method shows high precision (small deviation in between measurements) and improvable trueness (deviation from standard measurement method) to achieve accurate measurements. The results can be applied to further improvements of the method, anthropometric surveys, and CAD-integration.
Users
Please
log in to take part in the discussion (add own reviews or comments).