Daily physical activities recognition using wearable devices: a systematic mapping

Abstract

Sensors available in wearable devices can monitor their user's physical and mental health by recognizing the daily routine of physical activities. Because the area of human activity recognition can be complex due to a large amount of information, this work aims to facilitate the understanding of the area by providing a systematic mapping. A search using the Scopus database was conducted and resulted in 44 correspondences. The results were then classified according to their usage of primary techniques, preprocessing, feature extraction, sensors, datasets, most found activities, and considerations of lightweight algorithms. To better visualize the data, an online platform was also developed.