Combined range ambiguity resolution and noise reduction in lidar signal processing

Abstract

When the interval between transmitted pulses is shorter than the time of flight, it is not straightforward for a lidar to determine the distance uniquely. We present a method that uses varying pulse intervals, constructs a set of tentative point positions for each received pulse, and exploits the clustering of such tentative points to determine the correct distance with high probability. The clustering of true points also helps suppress noise pulses, even in a regime where the number of noise pulses is far greater than the number of true return pulses. It is then possible to use a detection threshold close to the noise level.