Body scan alignment reducing body posture variations for fit evaluation

摘要

This study aimed to provide a methodological solution for aligning 3D body scan images of torso posture variation in order to reduce inconsistency of the anthropometric data. Although a body scan system has been used for apparel applications such as fit evaluation, individual posture variations in the scanning process cause misalignment and inconsistency of the measurements. In this study, a new alignment technique, called Body Posture Variation Alignment (BPVA) aligns the source model to the target model to reduce posture variations. This BPVA technique has been tested on rotation determination, landmarks, and posture variations in flexion and extension. Aligning the two models at certain angles in the main slices significantly reduced the gaps. Overall, BPVA aligned the posture variations and increased a consistency of the anthropometric data. Results would be useful for the garment industry to provide accurate anthropometric measurements by aligning various body postures.