A Comparison of Women's Sizing by 3D Electronic Scanning and Traditional Anthropometry

Abstract

A collaborative project between University College London (UCL) and the London College of Fashion (LCF) resulted in a greater awareness of the possibilities and problems of using automated body-scanning techniques in measurement applications for clothing. In this study, 22 young females were measured, both by traditional hand methods and by the Hamamatsu Photonics 3D Body Line Scanner. The scanner uses infra-red sensors to sweep down the subject, providing a high-density 3D image of the whole body surface in one pass. Many practical issues arise in the large-scale use of electronic imaging systems, such as the choice of scan posture and automated extraction of accurate and reproducible measurements, without causing unnecessary discomfort to the subjects. The primary focus of this work was therefore to investigate a method suitable for large-scale surveying; to highlight specific problems of comparing electronic and manual anthropometric methods; and to present this information in the public domain. A total of 64 body measures were taken manually, and six scan poses were recorded for each subject. Software was used to extract measurements automatically from the 3D scans. Difference values between the electronic and manual measurements are presented and discussed. Results show good correspondence on many key measurements and higher errors on some others where particular problems occur in comparing electronic and manual measurements, such as those that are sensitive to posture or require the location of skeletal landmarks or features normally hidden by body tissue. This work was carried out as a preparatory study for a new UK collaborative initiative, the Centre for 3D Electronic Commerce. The primary goals of the Centre are to carry out a large-scale 3D survey of the UK population and to demonstrate the use of 3D body-scanning and virtual environments in custom clothing and on-line shopping.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 55 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.