Automation strategies and machine learning algorithms towards real-time identification of optically trapped particles

• Type: Article in International Scientific Journal
• Date: 2022

Authors

Oliveira, J (Author) Other; Rocha, V (Author) Other; Silva, NA (Author) Other; Jorge, PAS (Author) FCUP

Journal

Title: EPJ Web of Conferences

Vol. 266

ISSN: 2101-6275

Publisher: EDP Sciences

Other information

Authenticus ID: P-00X-SYC

DOI: 10.1051/epjconf/202226613025

Abstract (EN): <jats:p>To automatically trap, manipulate and probe physical properties of micron-sized particles is a step of paramount importance for the development of intelligent and integrated optomicrofluidic devices. In this work, we aim at implementing an automatic classifier of micro-particles immersed in a fluid based on the concept of optical tweezers. We describe the automation steps of an experimental setup together with the implemented classification models using the forward scattered signal. The results show satisfactory accuracy around 80% for the identification of the type and size of particles using signals of 250 milliseconds of duration, which paves the path for future improvements towards real-time analysis of the trapped specimens.</jats:p>

Language: English

Type (Professor's evaluation): Scientific