Systematic Error Reduction of i-TOF LiDARs Using Flexible Trapezoidal Waveforms

Abstract

Compared with common 3-D measurements technologies, indirect time-of-flight (i-TOF) systems offers significant advantages in volume, cost, power consumption, accuracy, range, and angular resolution. As such, they have found widespread applications in intelligent recognition, simultaneous localization and mapping (SLAM), and augmented reality (AR). However, due to the interference of systematic and random errors, current i-TOF systems achieve ranging accuracy only within several tens of millimeters. This severely limits their applications in high-precision scenarios, such as facial recognition payments, advanced manufacturing, and intelligent healthcare. In this work, we highlight that the most significant factor affecting accuracy among all errors is a systematic error known as wiggling. It is a high-dimensional complex function that nonlinearly couples with other systematic and random errors, making it difficult to independently separate, characterize, and compensate for. In light of this, we develop a methodology for ranging simulation and systematic error optimization based on adjustable trapezoidal functions derived from actual drive light waveform shaping. To demonstrate the effectiveness of the proposed theory and methodology, we perform measurement on an i-ToF system. Through global optimization of the frequency, duty ratio (DR), and rising/falling edge ratio (RFER) of the optical waveform, the total systematic error can be reduced from ±19 to ±4.5 mm under the conditions of a 5.6% RFER, a 33.2%, and a frequency of 100 MHz. By developing drive circuits that are optimized for the best DR and RFER, the systematic error is expected to be further reduced to the submillimeter level.