Direct time-of-flight (dToF) Lidar system based on single photon avalanche diode (SPAD) sensor plays a key role in the range finding applications because of its high sensitivity, low power consumption and high integrity. The travel time of photons backscattered from a distant target will be measured by the time-to digital-converters (TDC) to create a histogram whose peak location indicates the target distance. However, the bin width of histogram determined by the TDC’s time resolution (LSB) limits the precision of distance measurement. Therefore, sub-bin interpolation methods are introduced to facilitate super precision on-chip distance estimation. The balance between the performance and algorithm complexity must be kept very carefully, to maintain acceptable circuit area occupied, the power consumption and yet not to severely degrade performance. In this paper, we present the experimental results obtained from single point Lidar system composing of the Hangzhou Parity’s MPX102Q SPAD sensor as the receiver and a VCSEL pulse emitter as the transmitter. Several interpolation methods like the simple centroid or center-of-mass (CM) algorithm, and the gaussian mask centroid are applied to process raw histograms. The interpolation results of different algorithms are compared with the direct peak location method to examine their performance including distance resolution and ranging accuracy in different situations. Both sub-bin depth resolution, together with high precision and accuracy (mm level) covering a range up to 15m are achieved by this sensor containing integrated circuits for sub-bin interpolation, and this sensor is with potential to benefit the price-sensitive and power-constrained consumer products like sweeping robots.
|