3D Trajectory Reconstruction of High-Speed Irregularly Moving Objects via Single-Photon Spatiotemporal Correlation

Capturing high-speed moving objects, particularly those with irregular motion, has always been a fundamental pursuit in the field of imaging. However, limited by the frame rate, sensitivity, and temporal resolution, it has persistently posed a significant challenge. In this paper, a trajectory reconstruction scheme for high-speed irregularly moving objects using single-photon spatiotemporal correlation is proposed. Since the trajectories of moving objects are continuous in 3D space, the emitted or reflected photons exhibit spatiotemporal correlation. By leveraging the sparsity of correlation features, a compressed sensing algorithm for 3D trajectory reconstruction based on discrete random photon detection is developed. The spatial position and arrival time of photons are captured by a single-photon avalanche photodiode array. Here, deviating from the conventional approach, the trajectory is reconstructed directly by extracting the correlations of photons, the maximum trackable angular velocity up to 180 rad s⁻¹. Additionally, since background noise lacks spatiotemporal correlation, the proposed method exhibits excellent noise immunity. This ability to capture and track high-speed moving objects will provide new strategies for high-speed imaging applications in areas such as autonomous driving and the Internet of Things.