Wrist joint range of motion measurement based on Azure Kinect bone tracking
10.3969/j.issn.1006-9771.2022.08.014
- VernacularTitle:基于Azure Kinect骨骼追踪的腕关节活动度测量方法
- Author:
Wenbo ZHANG
1
;
Chang QU
1
;
Jianping ZHOU
2
;
Xiaoping ZHANG
1
;
Xiaotian ZHANG
1
Author Information
1. School of Mechanical Engineering, Nantong University, Nantong, Jiangsu 226019, China
2. Affiliate Hospital of Nantong University, Nantong, Jiangsu 226001, China
- Publication Type:Journal Article
- Keywords:
wrist;
range of motion;
Holt's double-parameter exponential smoothing method;
bone length constraint;
bone tracking;
Azure Kinect
- From:
Chinese Journal of Rehabilitation Theory and Practice
2022;28(8):981-988
- CountryChina
- Language:Chinese
-
Abstract:
ObjectiveTo automatically measure the wrist range of motion using the somatosensory capture device Azure Kinect. MethodsAzure Kinect was used to recognize the spatial coordinates of human elbow, wrist, palm, fingertip and other joint points, and the best measurement posture was determined by orthogonal experiment. Holt's double-parameter exponential smoothing method and bone length constraint method were used to smooth the joint point data to eliminate the jitter. The average angle of multiple frames was calculated through the spatial vector relationship to realize the automatic measurement of wrist joint range of motion. From May to October, 2021, 5 times × 10 groups of measurement were performed on each subject of wrist joint of five healthy subjects using the above method. ResultsThe R1
![]()
![]()
of each subject of wrist joint was between 83 and 127, and the value was in the middle. This method contrasted well with measurements made on augmented reality ruler, 2D images (r > 0.990), and the maximum error was within 1.61°. ConclusionThe wrist joint angle measurement modality using Azure Kinect is consistent to conventional measurement modality, and can meet the requirements of real-time measurement.
