Multi-camera orientation tracking method for anisotropic particles in particle-laden flows
Mees M. Flapper, Elian Bernard, Sander G. Huisman · 2025
Read Multi-camera orientation tracking method for anisotropic particles in particle-laden flows by Mees M. Flapper, Elian Bernard, Sander G. Huisman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-08694.txt.
Introduction
Mees Flapper, Elian Bernard, and Sander Huisman present a camera method that tracks not just where odd shaped particles are, but which way they point, in this 2025 paper. Rods, disks, and other non round grains tumble through fluids in ways spheres never do, and their turning matters for industry and nature alike. The method films particles from several viewpoints at high speed.
Knowing each particle's shape, the algorithm rebuilds its three dimensional position and orientation frame by frame. It follows many particles at once and can tell types apart. Tests on synthetic images map how noise, image size, and camera count affect accuracy. Real trials in still and turbulent fluids show the tracker working across very different shapes. The authors offer it as a general tool for flow studies.
Catalog record
| CREATOR | Mees M. Flapper, Elian Bernard, Sander G. Huisman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-08694
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Multi-camera orientation tracking method for anisotropic particles in particle-laden flows Mees M. Flapper, Elian Bernard, Sander G. Huisman Abstract A method for particle orientation tracking is developed and demonstrated specifically for anisotropic particles. Using (high-speed) multi-camera recordings of anisotropic particles from different viewpoints, we reconstruct the 3D location and orientation of these particles using their known shape. This paper describes an algorithm which tracks the location and orientation of multiple anisotropic particles over time, enabling detailed investigations of location, orientation, and rotation statistics. The robustness and error of this method is quantified, and we explore the effects of noise, image size, the number of used cameras, and the camera arrangement by applying the algorithm to synthetic images. We showcase several use-cases of this method in several experiments (in both quiescent and turbulent fluids), demonstrating the effectiveness and broad applicability of the described tracking method. The proposed method is shown to work for widely different particle shapes, successfully tracks multiple particles simultaneously, and the method can distinguish between different types of particles. I Introduction Tracking objects over time is vital for understanding dynamical systems and behaviours, ranging from pedestrian behaviour in crowds Corbetta et al. 2018 , sheep herd behaviour in a bottleneck Garcimartín et al. 2015 , to the transportation of particles on the water surface Salmon et al. 2023 . In fluid mechanics, particle tracking is a valuable instrument for investigating flows, particle-fluid interactions, and particle dynamics Schröder and Schanz 2022 . Tracking small tracer particles can be used to visualise complex dynamics in flows Godbersen et al. 2021 . Other fluid-dynamics applications of particle tracking are studying particle clustering dynamics Petersen, Baker, and Coletti 2019 , investigating the settling of snowflakes Nemes et al. 2017 , or measuring the dynamics of complex particle geometries Collins et al. 2021 . To most effectively use particle tracking as a measurement tool, the particle tracking method should ideally be able to track multiple (types of) particles simultaneously, should be flexible in camera arrangement (the relative location and orientation of the cameras), and track both the location and orientation of the particles. Preferably the tracking method can track anisotropic particles, since particles in general are anisoptropic, and isotropic particles are the exception. Hitherto scientific research has mainly focused on isotropic (spherical) particles Brown, Warhaft, and Voth 2009 ; Chiarini and Rosti 2024 , though recently some have studied anisotropic particles Voth and Soldati 2017 . These particles range from rigid fibres Bakhuis et al. 2019 and spheroids Will et al. 2021 , to flexible fibres and discs Verhille and Bartoli 2016 ; Verhille 2022 , hexagonal platelets Tinklenberg, Guala, and Coletti 2023 , and particles which break mirror symmetry Piumini et al. 2024 . Lagrangian tracking of these particles provides insights into particle dynamics, where both the particle’s location and orientation are relevant for studying the flow-particle and particle-particle interactions Toschi and Bodenschatz 2009 ; Brandt and Coletti 2022 ; Harms, Brunton, and McKeon 2024 . Many different tracking and measuring methods have been developed for obtaining the location and orientation data of spheres, fibres, and various anisotropic particles Cole et al. 2016 ; Mathai et al. 2016 ; Zimmermann et al. 2012 ; Ouellette, Xu, and Bodenschatz 2006 ; Ibarra, Bartoli, and Verhille 2023 . In this work, we describe a tracking method which finds the location and orientation of anisotropic particles using multiple cameras to image the particles. This method is flexible in terms of the camera setup, and works for particles with concave shapes, even those which have a centre of mass outside of the particle. Previous works have demonstrated 3D orientation tracking methods based on matching multiple 2D orientations Parsa et al. 2012 , finding a particle’s edges Brizzolara et al. 2021 , or a combination of the two Verhille and Bartoli 2016 , where multiple particles can be tracked simultaneously Bounoua, Bouchet, and Verhille 2018 ; Shaik et al. 2020 ; Baker and Coletti 2022 . More complex algorithms include methods using a particle marked with a black-and-white pattern Zimmermann et al. 2011 ; Mathai et al. 2016 ; Niggel et al. 2023 or projecting a synthetic particle model onto the experimental images Marcus et al. 2014 ; Cole et al. 2016 . This tracking method distinguishes itself from others by tracking multiple particle types simultaneously, being applicable to a wide range of particle shapes and sizes, including those that are concave and have their centre of mass outside the body. In addition, this method works for single or multiple cameras, and the camera arrangement is flexible. See Figure 1 , the shown particles can be tracked using this method (including particles of convex shapes), and the robustness of the method is carefully examined, such that it can be used as a tool for accurate location and orientation tracking. Figure 1: a) A photo of several 3D printed anisotropic particles tracked using the described method. From left to right, back to front: tetrad, oloid left-handed chiral particle, right-handed chiral particle. b) Synthetic 3D images of the used particles. From left to right: tetrad, left-handed chiral particle, right-handed chiral particle, oloid (not to scale). c) 3D sketch of the oloid geometry. The convex hull (‘shrink-wrapping’) of the two perpendicular disks gives the oloid surface. The grey lines span from one disk to the other, indicating the oloid surface. The working principle in this method is similar to previous methods using black-and-white particle projections, and relies on imaging a particle with one or multiple different directions. These images are then compared to calculated projections of a synthetic particle with a known orientation. The synthetic particle’s orientation is varied until the correct orientation is found. The details of this method are described step-by-step in section II . The robustness and accuracy of the method are quantified and evaluated in section III , detailing the effects of noise and image size using synthetic data. To close, section IV shows the usage of this method in test experiments, where we tracked chiral particles (particles breaking mirror symmetry, giving a left-handed and right-handed particle), tetrads, and oloids, all shown in Figure 1 . II Tracking principles A wide range of particle tracking algorithms have been developed, with different strong and weak points. Methods based on finding the 2D orientation or detecting the particle endpoints in 2D are very effective for rod-like particles Parsa et al. 2012 ; Bakhuis et al. 2019 ; Brizzolara et al. 2021 , but are not suited for more complex 3D shapes. A more general approach for 3D particle tracking is carving voxels based on multiple camera images, then finding the particle orientation by fitting the particle to the voxels Alipour, Paoli, and S. Ghaemi 2021 ; Giurgiu et al. 2024 . This method is shown to work well for tracking, and allows for a wide variety of geometries or configurations. Fu and Liu 2018 ; Masuk, Salibindla, and Ni 2019 . However, for our chiral particles, for specific camera configurations (for example when the cameras are closely arranged in a near-planar arrangement), we cannot distinguish the handedness of chiral the particle using this carving method. Therefore, for the chosen particle geometries, we do not have full camera flexibility