University of Twente Student Theses


Techniques to Process point cloud data to perform HAR using Machine learning

Anwar, Mian Tashfeen Shahid (2022) Techniques to Process point cloud data to perform HAR using Machine learning.

[img] PDF
Abstract:Human Activity Recognition (HAR) can solve a lot of problems for example it can be used to monitor elderly people in a better way and even in security purposes or to track movements in rooms. To perform Human Activity Recognition data in is collected in form of a point cloud using a (mmWave) millimeter wave radar. The problem we want to solve is that how can we do it effectively and accurately (accuracy , data , speed ) which is feasible for most use cases. In this paper first of all all methods for signal pre processing and Machine Learning (ML) were considered and then compared to shortlist the most important and effective ones and try to implement and test them. Our results are based on the accuracy achieved by the ML methods. We propose a new signal pre processing method which is a combination of frame divider , dbscan , rotations , enhanced voxelization and voxelization.
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