Interactive Measurement of Anthropometric Dimensions from Photo Image Based on Geodesic

Anthropometric data play a crucial role in ergonomics, and interactive measurement of anthropometric dimensions helps optimize design choices to enhance usability, comfort, and safety for users. Aiming at the problem that traditional body measurement methods, which lack real-time, remote and interactive measurement, we propose a novel anthropometric method based on geodesic and Fibermesh methods. First, a smooth three-dimensional human surface model is constructed from a two-dimensional photo using the Fibermesh method, adjusted according to weight and height ratios. Users can then interactively specify key measurement points on the 3D model based on their requirements. Next, the geodesic path between these points is calculated using a super-linear convergence algorithm, with the geodesic distance serving as the body dimension measurement. The experimental results show that our method achieves a total error of less than 5% in 3D body measurements, with a mean computational time of 1.85[Formula: see text]s for model reconstruction and an average measurement time of 0.026[Formula: see text]s. So that this approach enables real-time, remote anthropometry, supporting applications such as intelligent clothing customization.