User interaction and person recognition on smart TVs for elderly people

Abstract

Although technology is constantly developing at a rapid rate with newer and more innovative technologies being constantly put on the market, the different technological and digital techniques might be too advanced for older generations, making it increasingly difficult for them to adapt accordingly. Being somewhat unfamiliar with Smart TVs and its ancillary technological features, elderly people tend not to use its features and capabilities to their fullest potential. Unfortunately, even handling basic operations such as browsing the Internet or watching videos on YouTube and other platforms may prove to be tough and complicated to handle for the elderly. This thesis will address this problem by testing whether Gesture and Person Recognition can facilitate the interface between elderly people and Smart TVs or otherwise by comparing it to other types of user interactions. The latter includes a Remote Application and a Controller to see how the elderly people in the Casa Leone Elderly Home will react to such technologies. A prototype C+ application form using the OpenCV library was developed to tackle such problem. Thanks to the real-time functionalities of such library, the built-in camera was able to capture the elderly person‘s gestures and convert them into commands. These commands include different types of video categories which the participant will be able to watch depending on the Gesture input. The same procedure was repeated with the Remote application and the XBOX Controller to control the Smart TV by watching any video add-ons provided on the media application Kodi. Information was captured during the whole experiment to extract valuable data on which technology was preferred and how it can be improved in the future. From a total of forty participants that were asked to partake in this experiment, twenty participants were able to complete the whole study. The latter included the Pre-Study Questionnaire, a series of Tests with different Technologies and the Post-Study Questionnaire. The evaluation was conducted using the SPSS program to investigate and prove whether Gesture and Person Recognition are more reliable and easier to be implemented to control the Smart TV than other technologies.