AdaPT: Adaptable Particle Tracking for spherical microparticles in lab on chip systems
Description
Due to its rising importance in science and technology in recent years, particle tracking in videos presents itself as a tool for successfully acquiring new knowledge in the field of life sciences and physics. Accordingly, different particle tracking methods for various scenarios have been developed. In this article, we present a particle tracking application implemented in Python for, in particular, spherical magnetic particles, including superparamagnetic beads and Janus particles. In the following, we distinguish between two sub-steps in particle tracking, namely the localization of particles in single images and the linking of the extracted particle positions of the subsequent frames into trajectories. We provide an intensity-based localization technique to detect particles and two linking algorithms, which apply either frame-by-frame linking or linear assignment problem solving. Beyond that, we offer helpful tools to preprocess images automatically as well as estimate parameters required for the localization algorithm by utilizing machine learning. As an extra, we have implemented a technique to estimate the current spatial orientation of Janus particles within the x–y-plane. Our framework is readily extendable and easy-to-use as we offer a graphical user interface and a command-line tool. Various output options, such as data frames and videos, ensure further analysis that can be automated.