Validation study of a Kinect based body imaging system

2016

The purpose of this study was to introduce an automatic method for calculating personal body segment parameters (BSPs). In this automatic method, a Microsoft Kinect device was used to capture depth frames for measuring joint locations. The open source software, MakeHuman, was used for generating 3D human models by referring using the joint location data captured from the depth frames. Segmental meshes were obtained from the generated 3D human models and personal BSPs could be calculated automatically. The tests showed that the developed method can complete all of the processes without manual digitizing, anatomical landmark detection and medical scanner operation. Further research should be conducted to establish the accuracy of the segmental masses, centres of mass and moments of inertia acquired from the developed methods.

2016

Due to the importance of both developing and validating new low cost analysis, diagnosis and intervention devices for several areas, since rehabilitation until high performance training environments. The objective of this study is to determine the value for RMS error of the CM of body segments. Methods: anthropometric data were collected from a volunteer standing still in orthostatic position. Using Kinect™ for Microsoft™, the X, Y and Z coordinates were acquired. From there on, values of reference and values calculated by the device were taken into consideration. Results: the values calculated by the device demonstrated to be consisting through the different moments over data collects as well as RMS error values were satisfactory due to their low percentage, subsidizing favorably the usage of such device as a tool for analysis, diagnosis and intervention over several health care areas, such as physical education, physiotherapy and medicine among others.

Proceedings of 3DBODY.TECH 2022 - 13th International Conference and Exhibition on 3D Body Scanning and Processing Technologies, Lugano, Switzerland, 25-26 October 2022, 2022

3D scanning has evolved as one of the most advanced and accurate technology to measure humans and products. Quick and reliable results achieved by 3D scanning over manual measurements, make it the most preferred tool for measurement. 3D scanning has been extensively used in various national sizing surveys worldwide, including the INDIAsize (the National Sizing Survey of India) being carried out by the National Institute of Fashion Technology, INDIA. The 3D scan results are compared to the manual measurements to establish the accuracy of the scanner. This research paper describes two novel and alternative approaches (based on 1. Bias-shift, and 2. Regression modeling) to check the reliability and validity of the measures derived from 3D Body Scanner in comparison to measures provided by the manual measures. A comparison of both the approaches has also been discussed in this research paper. The 3-D Body scanning was done by Sizestream 3D Body Scanner-SS14. The manual body measurements were taken by experienced experts using an anthropometer, stadiometer, and certified flexible non-stretchable steel tape. In total 133 subjects (68 male and 65 female subjects) covering 102 body dimensions were taken manually and were used while comparing 3D scan measurements to establish the validity and reliability of the scanner. The procedure adopted for validation and reliability check for the 3D scanner was as prescribed in the ISO 20685(2005) and ISO 20685 (2018). It was observed that the Sizestream-SS14 scanners used were highly consistent in measuring the subjects as confirmed by the high values of Intra-class correlation coefficients (ICC) conducted to check for the consistency and repeatability between different scan measurements. However, a systematic error was reported in the process failing some of the measurements in terms of accuracy levels (as per ISO 8559 (1989) and ISO 20685(2005)) achieved against manual measurements used as the gold standards. Subsequently, two different approaches were applied to establish scanning accuracy and comparative analysis of results has been carried out. This research paper describes the validation and reliability procedure as per ISO protocols. It also discusses the regression-based statistical procedure adopted to confirm the desired measurement accuracy of the scanners within the permissible error limits of ISO 20685 (2005) and ISO 8559 (1989). Based on the comparative analysis, the paper also suggest recommended approach to achieve the desired accuracy by overcoming the systematic error in scanner measurement for all the anthropometric dimensions. This may help in making the data acceptable for use for any further analysis.

IOSR Journal of Computer Engineering, 2017

In this work we implemented a system to obtain human body parameter measurement without physically contacting the user. This implementation is contained the methods of obtaining 3D measurements using Kinect v2 depth sensor. The developed system at the initial stage is capable of detecting and obtaining personalized body parameters such as height, shoulder length, neck to hip length, hip to leg length and arm length by incorporating the necessary skeleton joints and front perimeter at chest, stomach and waist by incorporating the necessary 3D pixels. According to the results, the measurement on height and arm length of the person are relatively in good agreement with the actual values since the error is less than 5% and measurement has been taken in centimeters. Maximum 12% of an error incorporated of calculating front perimeter at chest. Experimental results obtained from the developed system are in acceptable range for dressing purpose and ultimately helpful for designing a real time 3D virtual dressing room.

Archives of Physiotherapy and Global Researches, 2016

From a cibernetic approach, the body system can be defined like a net of structural and functional related subsystems with motor equifinality inside the concept of balance, energetic economy and comfort: therefore the ideal posture is the one that allows the maximum effectiveness of motor gesture, in absence of pain with the maximum energetic economy. The present study research is based on the necessity to individuate a real objective evaluation system of the postural parameters, inexpensive and of simple use compared to the evaluation instruments in use already scientifically validated. The instrument used in this study is the Microsoft Kinect®, gaming platform combined with the Xbox console. Created by Microsoft in the field of play, the Microsoft Kinect® for years has entertained millions of consumers through the Motion Capture System, the recording of movement through cameras and instant or deferred replay. The primary aim of this randomized controlled single-blind research is the demonstration that, despite being commonly defined objective the evaluation systems that utilize markers, is essential to look for an alternative evaluation method to minimize the systematic human error. The results demonstrate the real validity of Kinect®, and have verified the reliability of the data obtained from the assessment, showing the scientific reliability of this innovative objective evaluation method in rehabilitation-clinical field.

2018

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 portab...

Proceedings of the 4th International Congress on Sport Sciences Research and Technology Support, 2016

Proceedings of 3DBODY.TECH 2018 - 9th International Conference and Exhibition on 3D Body Scanning and Processing Technologies, Lugano, Switzerland, 16-17 Oct. 2018, 2018

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.

Work (Reading, Mass.), 2012

Scan-derived landmarks locations and surface shapes are more and more used, but there is no commonly accepted protocol for evaluating the accuracy of these measurements. Therefore we propose a protocol for evaluating the accuracy of surface shape and the repeatability of scan-derived landmark locations. According to existing Japanese and German domestic standards, we propose to use an artefact (e.g. sphere with diameter of about 120 mm) calibrated very accurately for evaluating the accuracy of scanner-systems. For evaluating the repeatability of landmark locations, we propose to use an anthropomorphic dummy with landmark locations premarked. These test objects are measured by a 3D body scanner to be evaluated. Evaluation parameters such as trueness, precision, and repeatability are calculated from the measured data. A round-robin test was conducted in six different institutes using 17 body/head/foot scanners produced by eight companies. The purposes of the roundrobin test were to ev...

Analyzing human poses at workstations is a key issue in ergonomics in order to evaluate potentials risks of musculoskeletal disorders. Kinect looks promising in measuring 3D joint kinematics on site but few studies have estimated the accuracy of the kinematic data delivered by this sensor in real situations. Thus, this study aims at evaluating the accuracy of the Kinect sensor for work-related motions. To this end we compared ISB Euler joint angles estimated with a Kinect to those obtained with a wellestablished marker-based device. Several repetitions of four of the main manual tasks described in MTM-2 classification were performed. Results showed that the RMSE between the data obtained with the two systems were 11(±3)° and 25(±4)° for the shoulder and the elbow joints respectively. Moreover, the measurement error was not randomly distributed but strongly depended on the joint configuration. Better knowledge on the Kinect measurement errors should help scientists and engineers to design appropriate protocols to measure joint kinematics in ergonomic assessments with this promising device.