Fig. 1. Basic concept of MIR distance measurement based on asynchronous upconversion sampling. (a) The MIR probe and the pump beam are from two pulsed lasers, which are stabilized around a repetition rate fr with a slight frequency difference Δfr. The infrared probe is split by a beam splitter (BS) before being steered into reference and target surfaces. The reflections are then spatially mixed with the pump beam via a dichroic mirror (DM). The combined dual-color beams are sent into a nonlinear crystal to perform sum-frequency generation (SFG) in an asynchronous-pumping configuration. The resultant cross-correlation trace is measured by a silicon photodiode (PD), which allows for inferring the distance information through the time of flight. (b) The temporal separation between the signal and reference pulses is measured by using the nonlinear asynchronous optical sampling technique, where the MIR profile is precisely sampled by a pair-wise pump pulse at each period. Consequently, the cross-correlation trace is time stretched with a factor of M=fr/Δfr, which substantially alleviates the requirement for large detection bandwidth or low timing jitter in high-precision measurements.

Fig. 2. Experimental setup for the MIR single-photon upconversion ranging. The involved light sources are prepared from an Er-doped fiber laser (EDFL) and an Yb-doped fiber laser (YDFL2), which are stabilized at slightly different repetition rates fr1 and fr2 by referencing to a rubidium atomic clock (CLK). The output of EDFL at 1.56 μm is injected into a slave laser (YDFL1) at 1.03 μm to realize the passive optical synchronization. The synchronized dual-color pulses are injected into a periodically poled lithium niobate (PPLN) crystal to perform the difference frequency generation, which allows for the generation of MIR pulses at 3.1 μm. The MIR probe is split by a beam splitter (BS), and sent to a local reference and a distant target. The reflected light is then spatially combined with the amplified pump from YDFL2 via a dichroic mirror (DM2). The combined beam is steered into another PPLN crystal to perform an asynchronous upconversion sampling through the sum frequency generation. The asynchronous pulse pumping facilitates a fast and precise optical sampling on the MIR temporal profile, which results in a time-stretch cross-correlation trace at a wavelength of 0.77 μm. The distance information can be monitored in real time from the measured waveform detected by a low-bandwidth silicon photodiode (PD). Furthermore, MIR ranging at the single-photon level is implemented using the time-correlated single-photon counting (TCSPC) technique. The start channel is connected to the low-pass-filtered electric pulse from the SFG signal between the EDFL and YDFL2, while the stop channel is connected to the output of a single-photon counting module based on an avalanche photodiode (APD). Finally, the target distance can be measured from the accumulated photon histogram. YDFA: Yb-doped fiber amplifier; EDFA: Er-doped fiber amplifier; L: lens; M: mirror; FM: flip mirror; SPF, LPF, and BPF: short-pass, long-pass, and bandpass filters; NF: notch filter; Atten: optical attenuator; Col: fiber collimator; WDM: wavelength division multiplexer.

Fig. 3. High-resolution MIR ranging performance. (a) Recorded time-stretch waveform for the returned infrared probe from the reference and target. The update period Tupdate for each measurement is 1 ms, which corresponds to a ranging window about 46.45 ns. (b) Zoom-in of the measured peak for the target. The effective pulse duration of 9.8 ps is dictated by the cross-correlation width between the probe and pump pulses. (c) Measured distance as a function of the displacement for a target mounted on a translational stage. The initial position L0 is 46.45 cm and the travel range is 10 cm. (d) Residual and standard deviation at each measured position. (e) Allan deviation varies with different averaging time at a fixed target distance of L0. The minimum deviation of 5 μm is achieved for an averaging time about 0.5 s.

Fig. 4. High-speed MIR ranging for a rapidly moving voice-coil actuator. (a) Measured oscillating trace at a frequency of 15 Hz. (b) Involved frequency component is revealed by using a fast Fourier transform (FFT). (c) Recorded temporal trace when the actuator is driven by a synthesized voltage. (d) Multiple frequencies are identified via the Fourier analysis.

Fig. 5. High-sensitivity MIR ranging based on the time-correlated photon-counting technique. (a) Measured coincidence histogram for the returned infrared photons. Inset shows the zoom-in for recorded target peak. The peak width is about 0.213 μs, which corresponds to an effective time resolution of 9.9 ps. (b) Measured timing jitter of the used single-photon detector, which is the main limiting factor for ranging resolution in conventional ToF schemes. (c) Allan deviation measured at a detected optical energy of 8×10−5 photons/pulse. The minimum deviation of 4 μm is achieved for an averaging time about 64 s. (d), (e) Signal-to-noise ratio as a function of the detected power (d) and integration time (e). Insets show the measured histograms at representative conditions.

Fig. 6. Photon-counting MIR depth measurement for multi-interface structures. (a) Measured reference histogram in the absence of samples. (b)–(d) Measured histograms in the presence of silicon wafers with a thickness of L1 (b), a thickness of L2 (c), and both (d). The involved optical paths are indicated for the observed peaks in each case.