Optimization of System Design and Calibration Algorithm for SPAD-Based LiDAR Imager

Abstract

This paper presents the optimization of system design and calibration algorithm for LiDAR imager based on the single-photon avalanche diode (SPAD) image sensor. The supply voltage and the gating high-voltage driving circuit of the SPAD is optimized based on theoretical analysis and experimental data to improve the trigger times of the SPAD for the laser echo. The real-time calibration algorithm is realized in field programmable gate array (FPGA) to obtain the flight time and improve measurement reliability. A 10-bit time-efficient and area-efficient shifted inter-frame multi-event histogram and bubbling-based peak detection algorithm are used to calculate flight time, which requires the memory as low as 16 times smaller than storing complete histogram if the pixel values are coded on up to 10 bits and can detect up to four different targets simultaneously without increasing the measurement time. The performance of different kinds of background noise measurement methods is analysis based on experimental data, which shows that the measurement results are more accurate with the increasing of TDC event detection ability. The proposed crosstalk suppression algorithm based on LiDAR transmission function can reduce optical crosstalk between adjacent pixels to improve the measurement reliability of the LiDAR imager. The LiDAR imager is presented capable of measuring 10 m range with 70 Klux background noise and 30°×45° field of view.