Original scientific paper

https://doi.org/10.17559/TV-20240126001294

Edge Computing Assisted Internet of Things in Sports Management System

Baolei Zhang ; Nanfang College Guangzhou, Guangzhou, 510900, China
Juan Yang ; Shijiazhuang Preschool Teachers College, Shijiazhuang, 050228, China
Yan Peng ; North University of China, Taiyuan, 030500, China
Chong Liu ; Physical Education College, Chizhou University, Anhui, 247000, China *

* Corresponding author.

Abstract

In recent years, the movement analysis is commonly used to track the risk of injury and strengthen the efficiency of athlete performance. However, most of these devices are costly, found mainly in experimental settings, which analyze a few samples of each movement. In this paper, a new ambulatory movement analysis system with wearable sensors for the precise measurement of all athleteꞌs movements in an actual training scenario is introduced. Initially, an adaptive method categorizes a broad variety of training behaviors by the Differential Finite element Transformation method (DFET) along with a Random Forest Classification (DFET- RF) method. Secondly, the measurement of the absolute identities of the wearable sensor devices placed on the knee bone and pelvic bone is performed with a discrete gradient descent (DGD) algorithm, which calculates a range of motion-extension between the knee and hip angle. Finally, the edge computing is used to process data in real-time and reduce the latency of the system. The next version of wearable technology will know the person's identity, individually - not just physically and actively in a much more significant way; a wearable device that tells the world about the identity of the person and the connected devices. The knee flexion is greater at the terminal swing period (85%) and hip flexion (68%). The development of future device capabilities is based on verification. Once a wearable can validate the wearer's identity, several other things about their activities can be regulated. Such angles are automatically extracted for each movement during jogging at the acceleration of the sacrum effect. Besides, standard data has developed and is used to decide whether the movement methodology for a person varied from the standard data to classify potential instances due to injury. This is done by a gradient-shift recording technique for the joint-related angle details. For precise and automated assessment of athletic movements, effective activity measurement in various uncontrolled conditions for both injuries processing performance progress, the suggested system has been discussed in this paper.

Keywords

discrete transformation; discrete gradient; edge computing; finite element; random forest classification

Hrčak ID:

318491

URI

https://hrcak.srce.hr/318491

Publication date:

27.6.2024.

Visits: 406 *