The use of consumer depth cameras for calculating body segment parameters.

CLARKSON, Sean Alan. (2015). The use of consumer depth cameras for calculating body segment parameters. Doctoral, Sheffield Hallam University (United Kingdom)..

[img]
Preview
PDF (Version of Record)
10701291.pdf - Accepted Version
All rights reserved.

Download (21MB) | Preview

Abstract

Body segment parameters (BSPs) are pivotal to a number of key analyses within sports and healthcare. Accuracy is paramount, as investigations have shown small errors in BSPs to have significant impact upon subsequent analyses, particularly when analysing the dynamics of high acceleration movements. There are many techniques with which to estimate BSPs, however, the majority are complex, time consuming, and make large assumptions about the underlying structure of the human body, leading to considerable errors. Interest is increasingly turning towards obtaining person-specific BSPs from 3D scans, however, the majority of current scanning systems are expensive, complex, require skilled operators, and require lengthy post processing of the captured data. The purpose of this study was to develop a low cost 3D scanning system capable of estimating accurate and reliable person-specific segmental volume, forming a fundamental first step towards calculation of the full range of BSPs.A low cost 3D scanning system was developed, comprising four Microsoft Kinect RGB-D sensors, and capable of estimating person-specific segmental volume in a scanning operation taking less than one second. Individual sensors were calibrated prior to first use, overcoming inherent distortion of the 3D data. Scans from each of the sensors were aligned with one another via an initial extrinsic calibration process, producing 360° colour rendered 3D scans. A scanning protocol was developed, designed to limit movement due to postural sway and breathing throughout the scanning operation. Scans were post processed to remove discontinuities at edges, and parameters of interest calculated using a combination of manual digitisation and automated algorithms.The scanning system was validated using a series of geometric objects representative of human body segments, showing high reliability and systematic over estimation of scan-derived measurements. Scan-derived volumes of living human participants were also compared to those calculated using a typical geometric BSP model. Results showed close agreement, however, absolute differences could not be quantified owing to the lack of gold standard data. The study suggests the scanning system would be well received by practitioners, offering many advantages over current techniques. However, future work is required to further characterise the scanning system's absolute accuracy.

Item Type: Thesis (Doctoral)
Contributors:
Thesis advisor - Choppin, Simon [0000-0003-2111-7710]
Thesis advisor - Wheat, Jonathan [0000-0002-1107-6452]
Thesis advisor - Heller, Ben [0000-0003-0805-8170]
Additional Information: Thesis (Ph.D.)--Sheffield Hallam University (United Kingdom), 2015.
Research Institute, Centre or Group - Does NOT include content added after October 2018: Sheffield Hallam Doctoral Theses
Depositing User: EPrints Services
Date Deposited: 10 Apr 2018 17:23
Last Modified: 03 May 2023 02:06
URI: https://shura.shu.ac.uk/id/eprint/20644

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics