Body Sensor Network for Its Surface Shape Reconstruction
2023
Armands Ancāns

Defending
16.06.2023. 11:00, Rīgas Tehniskās universitātes Elektronikas un telekomunikāciju fakultāte, Āzenes iela 12, 201. auditorija.

Supervisor
Modris Greitāns

Reviewers
Aleksandrs Grakovskis, Dmitrijs Pikuļins, Mart Min

The thesis is dedicated to investigating novel methodologies for reconstructing body shapes using embedded sensors placed on the body. The aim is to develop a reliable and scalable system of body sensors capable of reconstructing stretchable and bendable shapes. The study encompasses a comprehensive review of existing approaches to body shape reconstruction and data acquisition from a network of body sensors. The research has led to the development of two significant innovations. Firstly, a novel technique for reconstructing the 3D coordinates of points that define the shape has been devised. This approach leverages orientation sensors embedded within elastic structures to reconstruct the coordinates of shape-defining points for objects experiencing stretching and bending deformations, without the need for supplementary sensors to measure alterations of distances between sensors. Numerical simulations have been conducted to validate the proposed method and establish a correlation between structural parameters and the accuracy of shape reconstruction. Furthermore, the simulations are utilized to compare the proposed approach to a conventional method that involves the placement of sensors directly on the body. The second innovation is an approach for data acquisition from a body sensor network with a large number of grouped sensors. This approach comprises an architecture for the optimized use of wires and a new communication protocol with minimized data overhead for low-power microcontroller interfaces in bus topologies. To validate the proposed methods in a laboratory setting, an experimental device with 26 orientation sensors was developed to reconstruct 12 points that define the shape of the arm. The reconstruction results obtained from this experimental device were then compared to those of the Optitrack optical marker tracking system. To evaluate the proposed data acquisition approach using the experimental device, the signals from the data bus were analyzed. Following the validation of the proposed methods through experiments and simulations, an evaluation of design choices was conducted, and recommendations were provided.


Keywords
Inerciālie sensori, virsmas rekonstrukcija, ķermeņu sensoru tīkli, valkājami sensori
DOI
10.7250/9789934229268

Ancāns, Armands. Body Sensor Network for Its Surface Shape Reconstruction. PhD Thesis. Rīga: [RTU], 2023. 78 p.

Publication language
Latvian (lv)
The Scientific Library of the Riga Technical University.
E-mail: uzzinas@rtu.lv; Phone: +371 28399196