Practical 3D human skeleton tracking based on multi-view and multi-Kinect fusion

Nguyễn, Mạnh Hùng; Hsiao, Ching-Chun; Cheng, Winton; Huang, Ching-Chun

doi:10.1007/s00530-021-00846-x

Cited by 14 publications

(10 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the depth camera of Kinect V2 solves the depth value according to the time‐of‐flight method, the diffuse reflection of the object surface from other infrared light sources will affect the solution of the depth value during the multi‐camera co‐sampling process. Considering the mutual interference between the instruments and the scope of the sensors, the angle between any two of the projections of the optical axes of the three cameras on the ground is 120°, the distance between each other is 4 m, and the three cameras are arranged in a regular triangle 7 …”

Section: Methodsmentioning

confidence: 99%

Sanda training program based on multi‐sensor data fusion in the internet of things environment

Zhou

2023

Internet Technology Letters

View full text Add to dashboard Cite

Sanda, also known as Chinese kickboxing, is a Wushu that combines elements of traditional Kung Fu, kickboxing, and grappling. Considering that the use of action recognition can effectively improve the level of Sanda players, this study uses inertial sensors and visual sensors to collect the motion data of Sanda. A federated Kalman filter with time‐varying parameters is designed according to the error form of each sensor to complete the data fusion. The main idea is that the whole system is divided into several local filtering systems, and a primary filter is designed for these systems. The main structure of the filter is a cascaded decentralized filtering structure composed of several local filters and a primary filter. The experimental results show that the proposed Sanda training scheme based on multi‐sensor data fusion under the Internet of Things environment has high action recognition accuracy, and the joint inertial‐vision model produces significantly better results than the other two models.

show abstract

Section: Methodsmentioning

confidence: 99%

Sanda training program based on multi‐sensor data fusion in the internet of things environment

Zhou

2023

Internet Technology Letters

View full text Add to dashboard Cite

show abstract

“…Kinect has its own algorithm to identify people and their respective 3D skeletons, but this system has problems when dealing with occlusions (either self or interpersonal ones) [5]. Recently, several works proposed approaches to estimate the 3D skeleton or 3D shape in a multiperson scenario, given 2D images.…”

Section: State Of the Artmentioning

confidence: 99%

Sharing Worlds: Design of a Real-Time Attention Classifier for Robotic Therapy of ASD Children

Santos¹,

Silva²,

Maddaloni³

et al. 2022

2022 International Conference on Rehabilitation Robotics (ICORR)

View full text Add to dashboard Cite

Robotic-assistive therapy has demonstrated very encouraging results for children with Autism. Accurate estimation of the child's pose is essential both for humanrobot interaction and for therapy assessment purposes. Nonintrusive methods are the sole viable option since these children are sensitive to touch. While depth cameras have been used extensively, existing methods face two major limitations: (i) they are usually trained with adult-only data and do not correctly estimate a child's pose, and (ii) they fail in scenarios with a high number of occlusions. Therefore, our goal was to develop a 3D pose estimator for children, by adapting an existing state-of-the-art 3D body modelling method and incorporating a linear regression model to fine-tune one of its inputs, thereby correcting the pose of children's 3D meshes. In controlled settings, our method has an error below 0.3m, which is considered acceptable for this kind of application and lower than current state-of-the-art methods. In real-world settings, the proposed model performs similarly to a Kinect depth camera and manages to successfully estimate the 3D body poses in a much higher number of frames.

show abstract

“…Human skeleton tracking systems often have difficulty handling lost tracking. Nguyen et al proposed a multiview system for 3D human skeleton tracking based on multicue fusion [5]. Mendicino et al aimed to develop and implement a complete integrated tracking system with very high accuracy both spatially and temporally per pixel [6].…”

Section: Related Workmentioning

confidence: 99%

Intelligent Optimization Algorithm of 3D Tracking Technology in Football Player Moving Image Analysis

Sun

Zhao

et al. 2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

Computer vision technology began to affect the development of football. There is increasingly high-tech in football broadcast technology, and many application tools have emerged in the field of football broadcast video analysis. The purpose of this paper is to study the improvement of target tracking algorithm for football broadcast video and to study the intelligent optimization algorithm of 3D tracking technology in football player moving image analysis. This paper proposes to select four models of YOLOv5 to perform target detection experiments in football broadcast videos and analyzes the principle of the Deep SORT multitarget tracking algorithm. At the same time, it is based on the 3D tracking and 3D pose estimation of players based on cross-view correlation matching, and to measure the comprehensive performance of the tracker in the football scene, experiments are carried out on the accuracy and speed of the tracker under the football datasets of four different scenes. The experimental results in this paper show that the MOTA values corresponding to the 3D tracking results and 2D projection results obtained in the campus dataset are only 50 and 56.2. This is much lower than the tracking performance when based on other similarity matrices. The MOTA value of the obtained tracking result (92.6) is very close and significantly outperforms other methods. CCOT performs better on datasets 28, 29, 31, and 32, ECO stands out on dataset 38, and Siamese also performs well on datasets 22 and 36.

show abstract

Practical 3D human skeleton tracking based on multi-view and multi-Kinect fusion

Cited by 14 publications

References 37 publications

Sanda training program based on multi‐sensor data fusion in the internet of things environment

Sanda training program based on multi‐sensor data fusion in the internet of things environment

Sharing Worlds: Design of a Real-Time Attention Classifier for Robotic Therapy of ASD Children

Intelligent Optimization Algorithm of 3D Tracking Technology in Football Player Moving Image Analysis

Contact Info

Product

Resources

About