Abstract:
The ability to capture the spatial dimensions of the world around us is growing in importance, with the widespread adoption of 3D-sensing used today for secure facial authentication, AR occlusion, robotic vision and SLAM, autonomous driving, and 3D-reconstruction. Most state-of-the-art light detection and ranging (LiDAR) systems mainly focus on the sensor design [1]–[4]. However, the optical-electrical system of LiDAR is complex, requiring hardware and software co-optimization across the entire signal chain: high-power sub-1ns pulsed laser drivers, high-efficiency lasers, class-1 laser eye-safety, optical lens for focusing or diffusion, high-SNR single-photon detection receiver arrays, and machine learning (ML) based computational photography.