Solid-State Dtof LiDAR System Using an Eight-Channel Addressable, 20w/ch Transmitter, and a 128x128 SPAD Receiver with SNR-Based Pixel Binning and Resolution Upscaling.
Shenglong Zhuo,Lei Zhao,Tao Xia,Lei Wang,Shi,Yifan Wu,Chang Liu,Chill Wang,Yuwei Wang,Yuan Li,Hengwei Yu,Jiqing Xu,Aaron Wang,Zhihong Lin,Yun Chen,Rui Bai,Xuefeng Chen,Patrick Yin Chiang
DOI: https://doi.org/10.1109/cicc53496.2022.9772823
2022-01-01
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.