3D Human Models from 1D, 2D and 3D Inputs: Reliability and Compatibility of Body Measurements

This paper presents partial results of a larger validation study of different Data-driven 3D Reconstruction (D3DR) technologies developed by IBV to create watertight 3D human models from measurements (1D3D), 2D images (2D3D) or raw scans (3D3D). This study quantifies the reliability (Standard Error of Measurement, SEM; Mean Absolute Deviation, MAD; Intra-class Correlation Coefficient, ICC; and Coefficient of Variation, CV) of body measurements taken on human subjects. Our results are also compared to similar studies found in literature assessing the reliability of digital and traditional anthropometry. Moreover, we assess the compatibility (bias and Mean Absolute Error, MAE) of measurements between D3DR technologies. The results show that 2D3D can provide visually accurate body shapes and, for the measurements assessed, 2D3D is as reliable as high resolution 3D scanners. It is also more accurate than manual measurements taken by untrained users. Due to accessibility, cost and portability (e.g. 2D3D built in a smartphone app) they could be more suitable than other methods at locations where body scanners are not available such as homes, medical or physical therapy offices, and small retail stores and gyms.