1. Principles of 3D Reconstruction: Multiple Viewpoints + Triangulation

A single camera cannot capture the motion trajectory of an individual particle within a flow field in three-dimensional space. Only by fusing observations from multiple viewpoints (typically using 3–4 cameras) can the particle’s position be precisely localized in space. The cameras are arranged in a photogrammetric configuration, observing the same spatial region from different viewpoints with slightly overlapping fields of view. Based on corresponding pixels and known camera positions, the three-dimensional coordinates of the particle in space can be calculated using triangulation. The more observation angles used, the more accurate and stable the reconstruction of the particle’s trajectory becomes.

2. Key Feature: Precise Camera Synchronization

Precise synchronization is key to the successful fusion of multi-camera data in subsequent stages. The U-See partner used in the experiment IDS Event Camera, offers two extremely useful interface features:

• Trigger Input: Assign a unique "timestamp" to all data streams to ensure that subsequent events can be accurately matched.

• Hardware Synchronization (TDRSTN): The camera can be set to start automatically even when connected to a different computer.

3. Data Processing: From Events to Movement Trajectories

Once data collection is complete, the real challenge begins: first, the event data from each camera must undergo geometric registration (i.e., camera calibration). Subsequently, the particles are spatially localized using two methods:

• Navigate directly from the sync event;

• Alternatively, a two-step approach can be used: first, track the particles from a single viewpoint, and then reconstruct their trajectories.

Accumulate the positions and timestamps of individual pixel events over a specified time interval—that is, fuse them in both time and space—to ultimately form "trajectories" in space, intuitively illustrating the motion of particles within a spatial volume over time. This form of visualization is crucial for understanding complex flow fields: it allows for clear observation of particle trajectories, the presence of turbulence, and how shock waves propagate. Phenomena such as backflow, vortex formation, and local velocity changes can also be depicted in this manner. Such qualitative visualizations are not only highly valuable for scientific research and education but also widely contribute to the development and optimization of industrial systems in cutting-edge fields such as aerospace, fluid mechanics, and microfluidics.