HIGIA - Sistema de reconocimiento de actividades higiénicas con sensores multimodales
Fecha
2022-01-10
Autores
Título de la revista
ISSN de la revista
Título del volumen
Editor
Jaén: Universidad de Jaén
Resumen
En este Trabajo de Fin de Máster se desarrolla un sistema de sensorización y reconocimiento de actividades centrado en los
hábitos de higiene de usuarios en espacios inteligentes. En concreto, se ha propuesto la sensorización del cuarto de baño y las
actividades que se producen en el mismo, clave en las necesidades básicas humanas (correcta eliminación y hábitos de higiene).
Para ello, se han integrado dispositivos multimodales de diferente naturaleza de acuerdo a los requisitos de no invasividad en
condiciones naturalistas: sensores de temperatura y humedad para detectar duchas, sensores vestibles para detectar diferentes
eventos relacionados con la higiene (lavarse los dientes, lavarse las manos, etc), sensores de gas en el aseo para detectar las
deposiciones, sensores de pH, ORP y temperatura en los aseos para detectar las excreciones humanas, sensores de audio para
predecir eventos asociados a sonidos y sensores de visión termal para estimar la pose de una persona en la estancia.
This Master's Thesis develops a sensorisation and activity recognition system focused on the hygiene habits of users in smart environments. Specifically, we have proposed the sensorisation of the bathroom and the activities that take place in it, which are key to basic human needs (correct elimination and hygiene habits). To this end, multimodal devices of different nature have been integrated according to the requirements of non-invasiveness in naturalistic conditions: temperature and humidity sensors to detect showers, wearable sensors to detect different events related to hygiene (brushing teeth, washing hands, etc.), gas sensors in the toilet to detect depositions, pH, ORP and temperature sensors in the toilets to detect human excretions, audio sensors to predict events associated with sounds and thermal vision sensors to estimate the pose of a person in the room.
This Master's Thesis develops a sensorisation and activity recognition system focused on the hygiene habits of users in smart environments. Specifically, we have proposed the sensorisation of the bathroom and the activities that take place in it, which are key to basic human needs (correct elimination and hygiene habits). To this end, multimodal devices of different nature have been integrated according to the requirements of non-invasiveness in naturalistic conditions: temperature and humidity sensors to detect showers, wearable sensors to detect different events related to hygiene (brushing teeth, washing hands, etc.), gas sensors in the toilet to detect depositions, pH, ORP and temperature sensors in the toilets to detect human excretions, audio sensors to predict events associated with sounds and thermal vision sensors to estimate the pose of a person in the room.