University of Twente Student Theses


Human activity recognition using two millimeter-wave radars

Eshy, A. (2022) Human activity recognition using two millimeter-wave radars.

[img] PDF
Abstract:Human activity recognition (HAR) aims to label, recognize, track human activities accurately, and it has been implemented through several approaches, such as ambient sensors, cameras, or wearable devices. However, in privacy-sensitive areas, a camera could collect extraneous ambient information that a user may not feel restful revealing. Therefore, millimeter wave (mmWave) radars have been proposed as an alternative for detecting and tracking human activity. The mmWave radars endure the unique advantage of being effective under non-line-of-sight scenarios, effectively capture a minimal subset of the ambient information using micro-Doppler spectrograms producing higher accuracy, and can track the user while preserving privacy. The article proposes an approach that can detect human activity recognition and track the human user accurately by using two-millimeter wave (mmWave) radars. The approach focuses on advanced machine learning algorithms, innovations in hardware architecture, and decreasing monitoring costs. This paper proposes RadHAR, a framework that performs accurate human activity detection using point clouds. The collected human activity data-set got evaluated, and a comparison of the accuracy of various classifiers on the data set found that the best-performing deep learning classifier achieves an accuracy of 97.71\%. The evaluation shows the efficacy of using two mmWave radars for accurate HAR detection and reliable tracking.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page