Artificial vision system for the prevention of injuries in the upper back and neck areas based on the OpenPose algorithm
| Autor: | Heraclio García-Cervantes, Juan Carlos Cardona-Villalpando, Alan David Blanco-Miranda, Didia Carrillo-Hernández |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
03 medical and health sciences
0302 clinical medicine 020205 medical informatics business.industry Computer science 0202 electrical engineering electronic engineering information engineering Computer vision 02 engineering and technology Artificial intelligence business Artificial vision system 030217 neurology & neurosurgery |
| Zdroj: | ECORFAN Journal Republic of Paraguay. :13-19 |
| ISSN: | 2414-4827 |
| DOI: | 10.35429/10.35429/ejt.2020.8.4.13.19 |
| Popis: | Computer vision is a tool used to understand images that have been analyzed and processed. And in health and human body applications it is no exception, since they help to vectorize the body and its movement, analyzing the changes between them. The main objective of this work is to develop an interactive mobile application for the monitoring, description and recording of body postures in the neck and upper back areas when using cellular devices. The OpenPose algorithm is used to identify and register the specific points of the established zones. In the analysis process, the Keras framework is used in order to build a deep learning convolutional neural network. A graphical user interface is designed in order to facilitate the use and interpretation of measurements. In addition, it is accompanied by an alert system sent when the maximum load angle accepted by the neck in an adult is exceeded (30 ° with respect to the imaginary vertical line on the back). This work contributes to the prevention of future injuries in the cervical areas according to the information provided by specialists in physical rehabilitation. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|