Activity recognition in intelligent assistive environments through video analysis with body-angles algorithm a first step for future behaviour recognition

Abstract

Activity Recognition in a scientific setting is a field that is extremely popular, in which numerous and diverse proposals exist that tackle the creation of systems capable of recognising activities through different types of sensors. Given the relative maturity of Activity Recognition in comparison to Behaviour Recognition, most of the existing proposals in this last field are based in Activity Recognition but with the difference of analysing the activities throughout time. Therefore, the objective of this article is to describe the first phases of development of a larger scale research (doctoral thesis) with which we will intend to analyse the Behaviour of people with focus not only based on Activity Recognition but also with a strong component centered around smart environments with context awareness and supported by the foundations of The Psychology of behaviour.