Reconeixement d'accions mitjançant visió artificial per teleassistència

The continued aging of the population has made in recent years gives rise to special interest research and development of systems that help increase the quality of life of older persons without involving an unsustainable spending by the social security system. In this context, telecare has experienced some improvements incorporating sensors that detect potential health problems or accidents, thus extending the stay in the home and avoiding or delaying an entry into a residence. This work explores the possibility of using artificial vision to monitor actions of everyday life through the imaging method of the history of the movement (MHI) and the classification of these images using neural networks and support vector machine. The results obtained are insufficient to move the real system built on a solution, but show the possibilities that this method can bring. A possible improvement of the results could be given by expanding the data set labeled or improving image processing. Another possibility for improvement could come by the combination of the method with other artificial vision and the detection and tracking of body parts.

Grantor dc:publisher

Universitat Oberta de Catalunya

Year dc:date.issued

2016

Author dc:creator

• Sánchez Pérez, Samuel

• teleassistència
• visió artificial
• aprenentatge automàtic
• teleasistencia
• visión artificial
• aprendizaje automático
• telecare
• artificial vision
• automatic learning

Licence dc:rights.uri

• http://creativecommons.org/licenses/by-nc-nd/3.0/es/

Language dc:language.iso

cat

Handle dc:identifier.uri

https://hdl.handle.net/10609/52325

OAI identifier oai:identifier

oai:openaccess.uoc.edu:10609/52325