Manasi Muglikar, Guillermo Gallego and Davide Scaramuzza

Event cameras are bio-inspired sensors providing significant advantages over standard cameras such as low-latency, high temporal resolution, and high dynamic range. We propose a novel structured-light system using an event camera, to tackle the problem of accurate and high speed depth sensing. The setup presented in this paper consists of an event camera and a laser-point projector, which uniformly illuminates the scene in a raster scanning pattern in 16 milliseconds. Previously proposed methods rely on time-constancy principle (i.e., instantaneous transmission from projector to camera), which often breaks down due to signal latency and jitter effects, leading to noisy correspondences and depth errors at high scanning speeds. Our method consists of minimizing an energy function designed to exploit the spatio-temporal consistency of light rays projected from the projector plane and observed as events on the event image plane. The resulting method is robust to event jitter and can therefore perform better at higher scanning speeds. Experiments demonstrate that our method can deal with high-speed motion and outperform state-of-the-art 3D reconstruction methods based on event cameras, reducing the fill rate by 83% on average, for the same acquisition time. Code and dataset will be released upon acceptance.