Invited Paper
Medium Altitude Airborne Geiger-mode Mapping Lidar System
William E. Clifton*a, Bradley Steeleb, Graham Nelsonb, Antony Truscottc, Mark Itzlerd, Mark
Entwistled; aHarris Corporation, 1025 W NASA Blvd., Melbourne, FL USA 32919; bQuartus
Engineering, Inc., 10251 Vista Sorrento Parkway, STE 250, San Diego, CA USA 92121; cTruscott
Designs, San Diego, CA USA; dPrinceton Lightwave, Inc., 2555 Route 130 South, Cranbury, NJ
USA 08512
ABSTRACT
Over the past 15 years the Massachusetts Institute of Technology, Lincoln Laboratory (MIT/LL), Defense Advanced
Research Projects Agency (DARPA) and private industry have been developing airborne LiDAR systems based on
arrays of Geiger-mode Avalanche Photodiode (GmAPD) detectors capable of detecting a single photon. The extreme
sensitivity of GmAPD detectors allows operation of LiDAR sensors at unprecedented altitudes and area collection rates
in excess of 1,000 km2/hr. Up until now the primary emphasis of this technology has been limited to defense
applications despite the significant benefits of applying this technology to non-military uses such as mapping,
monitoring critical infrastructure and disaster relief. This paper briefly describes the operation of GmAPDs, design and
operation of a Geiger-mode LiDAR, a comparison of Geiger-mode and traditional linear mode LiDARs, and a
description of the first commercial Geiger-mode LiDAR system, the IntelliEarth™ Geospatial Solutions Geiger-mode
LiDAR sensor.
Keywords: LiDAR, LADAR, Geiger-mode, avalanche photodiode, single photon, mapping, FOPEN
1. INTRODUCTION
LiDAR systems based on avalanche photodiodes operating in Geiger-mode have been under continuous development
since the late 1990’s at the Massachusetts Institute of Technology, Lincoln Laboratory (MIT/LL) [1,2,3,4,5]. Since then,
several airborne Geiger-mode LiDAR systems have been developed: JIGSAW, a low altitude foliage penetration
(FOPEN) LiDAR shown in Figure 1 [6,7], the Airborne Lidar Research Testbed (ALIRT) a medium altitude mapping
LiDAR [8], the High Altitude Lidar Operations Experiment (HALOE) a high altitude imaging LiDAR [9] and the
MIT/LL MACHETE a medium altitude FOPEN LiDAR system. DARPA facilitated the transfer of GmAPD technology
from MIT/LL to Princeton Lightwave Inc. [10, 11], and Boeing Spectrolab [12], in order to establish commercial sources
of GmAPD based LiDAR cameras.
During the same period, compact, neodymium-doped yttrium aluminum
garnet (Nd:YAG) Diode Pumped Solid State Lasers (DPSSLs) with high
pulse energies, high pulse repetition rates (PRFs) and subnanosecond pulse
widths were also under development. The maturation of these two
complementary subsystems has made a commercial GmAPD LiDAR capable
of collecting high resolution LiDAR data at area coverage rates (ACRs)
exceeding 1,000 square kilometers per hour technically and economically
feasible.
This paper briefly describes the operation of GmAPDs, the basic design and
operation of a Geiger-mode LiDAR, a comparison of Geiger-mode and
traditional linear mode LiDARs, and a description of the first commercial
Geiger-mode LiDAR system, the IntelliEarth™ Geospatial Solutions Geigermode LiDAR sensor (Harris Geiger-mode LiDAR).
Figure 1. Jigsaw FOPEN Geiger
mode LiDAR System (2005)
william.clifton@harris.com; tel: 1.321.693.4873; http://govcomm.harris.com/solutions/products/isr/geoint_lidar.asp
Laser Radar Technology and Applications XX; and Atmospheric Propagation XII, edited by
Monte D. Turner, Linda M. Wasiczko Thomas, Gary W. Kamerman, Earl J. Spillar, Proc. of SPIE
Vol. 9465, 946506 · © 2015 SPIE · CCC code: 0277-786X/15/$18 · doi: 10.1117/12.2193827
Proc. of SPIE Vol. 9465 946506-1
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
2. GEIGER-MODE AVALANCHE PHOTODIODE LIDAR
Geiger-mode Avalanche Photodiode Detector Operation
In linear mode, when the reverse-bias
voltage
across
an
avalanche
photodiode is maintained below a
critical level the output current is
linearly proportional to the incident
photon flux. Above this critical level,
or avalanche breakdown voltage, the
absorption of a single photon can
create an avalanche current pulse
large enough to be detectable by
digital timing circuitry. When
Digital
operated with a reverse-bias voltage
Timing
above the avalanche breakdown
Circuit
voltage, the APD is in Geiger-mode.
Thousands of GmAPDs can be
Figure 2. Geiger mode Avalanche Photodiode focal plane array
integrated into focal plane arrays
(FPAs) with integral time-of-flight
(TOF) counters and associated readout integrated circuitry (ROIC). In addition, because the active area of the detector is
relatively small compared to the spacing between adjacent detectors, a microlens array with a dedicated lens for each
detector is added to increase the effective fill factor to ~75%. This maximizes the probability for any photon incident on
the FPA to be detected (Figure 2)
Flash
Detect
H
arm (glee ormilae)
arm delay
Y
sal- -range gate Eel
-
Gege
range gale duration
disarm (remove over -Was)
Eisarm del y
The operation of a GmAPD LiDAR
is conceptually very simple. A laser
pulse is used to illuminate a scene.
By precisely measuring the TOF of
the photons reflected off the scene
the three dimensional coordinates of
features within the scene can be
determined.
Figure 3 shows the timing sequence
of a GmAPD imaging cycle. T0 is
defined by when a transmit laser flash is detected. Increasing the bias on the array (arming) is delayed based on a priori
knowledge of the approximate distance to the target. Following the arm delay, the capacitance of each GmAPD is
charged up to the desired overbias voltage, determining the effective sensitivity or Photon Detection Efficiency (PDE) of
the detectors in the array. Once the pixels are armed, the voltage source is disconnected, and any subsequent avalanche
event simply discharges the diode capacitance, with the associated current flow limited to just the amount of charge
needed to remove the overbias from the detector.
Figure 3. Geiger mode Lidar imaging cycle
Once the array has been over-biased and operating in Geiger-mode, the range gate is opened by starting the ROIC TOF
counters. The range gate duration is chosen to completely encompass the volume of interest. While the range gate is
open, photons incident on a detector may be converted into a photoelectron, which in turn may cause an avalanche,
latching the TOF counter of the detector. Once the gate is closed, the over bias is removed, the array TOF values are read
out and the array is then ready for another cycle.
Proc. of SPIE Vol. 9465 946506-2
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
Key Characteristics of GmAPDs
•
•
•
•
•
•
•
•
Single-photon sensitivity.
o Reduces system size, weight and power
o Allows operation at high altitude
Compact detection circuitry makes fabrication of large arrays of detectors practical.
Large avalanche current flow with sharp leading edge is detectable by simple CMOS digital timing circuitry.
Noiseless readout using digital detection
Low timing jitter of < 500 ps
Large avalanche current flow causes “after-pulse” avalanches due to carrier trapping and de-trapping if the array is
re-armed before the de-trapping process has completed which limits the detection rate of any given pixel [13].
o Since the duration of the de-trapping process is on the order of hundreds of nanoseconds, compared to the
maximum frame rate of current GmAPD cameras which is on the order of hundreds of kilohertz, after-pulsing
has no impact on performance of the system.
Hot carrier luminescence during avalanches can cause photon emission, resulting in optical crosstalk between
detectors.
o Crosstalk can be reduced by FPA designs incorporating trenches between detectors, reduced detector active
area diameters, antireflective coatings, as well as reducing the sensitivity of the FPA and post collection
processing.
A GmAPD detector as currently implemented has only one measurement opportunity per imaging cycle, once the
detector has avalanched it cannot detect any subsequent photons until the current imaging cycle is completed and
the detector is rearmed. The time between the avalanche and when the detector can again detect an incoming photon
is referred to as blanking loss
o Compensated for by high PRF lasers and large detector arrays, enabling sample rates in excess of 400 MS/s
o Similar to after-pulsing, the current and projected synchronous frame rates of GmAPD FPAs and
repetition rates of short pulse high energy laser transmitters, blanking loss currently has no impact on the
performance of a direct detection TOF based system.
o Blanking loss is not an inherent limitation of GmAPDs. Future asynchronous ROICs will enable multiple
TOF measurements per imaging cycle.
Basic Measures of Performance of Geiger-mode Avalanche Photodiodes
Two primary measures of performance of a GmAPD detector are the Photon Detection Efficiency (PDE) and the Dark
Count Rate (DCR). [14]
PDE is the probability that an incident photon causes the detector to fire and is a combination of the quantum efficiency
of the detector and the probability that an avalanche occurs when a single photon reaches the detector. However, if the
average number of photons per pulse is greater than one, the effective Multi-photon PDE (PuDE) increases significantly.
For example, for a GmAPD FPA with a single photon PDE of 30% receives an average of ten photons per pulse the
PuDE increases to 98%. [10]
DCR is the probability that a detector fires in the absence of a photon and is expressed as counts per second per detector
and expressed in kHz. At a PDE of 30% a typical DCR is on the
Time of Flight
order of 20 kHz so during a 4 microsecond 600 meter range gate
an average of 0.08 noise photoelectrons will be randomly
distributed through the range gate for each detector. DCR has
significantly less impact than solar background and can be
ignored in most daytime radiometric analyses.
Geiger-mode and Linear mode LiDAR systems
500
1000
1500
2000
2500
Geiger-mode and linear mode LiDARs are two different methods
of accomplishing the same task, which is to reconstruct the TOF
waveform returned from the scene of interest. While a linear
time of flight bin
Figure 4. Geiger mode Lidar range histogram
Proc. of SPIE Vol. 9465 946506-3
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
mode system samples the entire waveform in a single flash, a Geiger-mode system samples the waveform one flash at a
time at a very high sample rate. For example, the waveform shown in Figure 4 is the number of counts per range bin
collected over multiple flashes of a GmAPD LiDAR. The GmAPD LiDAR samples the waveform one point at a time, so
unlike a linear mode LiDAR, does not require a high speed digitizer which results in a simpler, highly scalable and more
cost effective implementation.
In medium and high altitude applications a GmAPD LiDAR has a significant advantage over traditional linear mode
linear mode systems due to the single photon sensitivity of the detectors. Compared to a GmAPD LiDAR, a traditional
linear mode system requires at least an order of magnitude more photons for a detection which in turn drives up the
average power requirements and overall size, weight and power (SWaP) of the system making a medium or high altitude
linear mode LiDAR impractical.
°s:8
!
I
I
!
r
äì {. .w,,
JM..._
` e"eeiq
.e,.e.,,e,.í!:.
__. wwR
i!!!e!w!!
40411
2
Figure 5. Geiger mode APD 2D relative intensity image
Another difference between a GmAPD LiDAR and linear mode LiDAR is the method of building an intensity image.
The linear mode LiDAR is capable of creating an intensity image in a single flash, while a GmAPD requires multiple
flashes to build up an intensity image of the scene which may be illuminated by the transmit laser or solar background as
shown in Figure 5.
3. INTELLIEARTH™ GMAPD LIDAR SENSOR
Sensor Hardware
Figure 6. Harris IntelliEarth™ Geospatial Solutions Geiger-mode LiDAR sensor
on alignment stand in the optical laboratory
Proc. of SPIE Vol. 9465 946506-4
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
The Harris IntelliEarth™ Geospatial Solutions Geiger-mode LiDAR sensor is the first commercial airborne LiDAR
system that takes advantage of the single photon capabilities of the Geiger-mode avalanche photodiode (Figure 6).
The primary components of the system (Figure 7):
•
•
•
•
•
•
•
•
•
•
•
128x32 InP/InGaAsP Geiger-mode camera capable of readout rates in excess of 100 kHz (1)
Compact Nd:YAG diode pumped solid state laser (2)
270 mm Holographic Optical Element (HOE) scanner with 15° scan half angle capable of rotational speeds in
excess of 2,000 RPM. (US Patent US 2015/0029571 A1 January 29, 2015) (3)
Real time transmit LOS adjustment relative to receive LOS which compensates for scanner motion during pulse
round trip time.
High efficiency narrow bandpass filter reduces solar background noise.
Transmit beam shaping optic optimizes illumination pattern on the ground.
Nadir looking Ritchey – Chrétien telescope for collection of returned light
Inertial navigation system (INS/GPS) including Inertial measurement unit (IMU) (4 and 5) provides precise
LOS knowledge.
High speed flash detector for precision laser pulse timing
Data acquisition electronics (6)
Sensor controller
Figure 7. Sensor configuration and major subsystems
In operation, when the laser (2) fires, a flash detector triggers the TOF timers within the camera (1). The pulse,
meanwhile, is deflected through the central transmit prism which has been adjusted (i.e., “clocked) to compensate for the
motion of receive field of view during the two way transit time of the laser pulse. The return pulse is collected by the
Palmer scanner (3), passes through the narrow bandpass filter to reduce the amount of solar background, then to the
128x32 GmAPD camera (1). The measured TOF and metadata values are then passed to the processing subsystem via
the data acquisition subsystem (6) which combines the camera TOF data with all the metadata required for creating point
cloud data as well as system health and status data. The associated line of sight orientation and sensor position data is
determined by the IMU (4) via the INS/GPS (5).
Given the ability of a GmAPD detector to measure a single photon, minimizing solar background is an important
consideration during the design of the system. The choice of system wavelength has a significant impact on the solar
background level and overall system efficiency. For example, a LiDAR using silicon based detectors with a frequency
doubled Nd:YAG laser operating at a wavelength of 532 nm is subject to three times the solar irradiance when compared
to a Nd:YAG LiDAR system operating at a wavelength of 1064 nm. In addition, the 1064 nm Nd:YAG laser is
Proc. of SPIE Vol. 9465 946506-5
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
approximately twice as efficient as the frequency doubled Nd:YAG laser. Other effective methods of reducing solar
background is minimizing the system aperture, installation of a narrow bandpass filter in the receive path, reducing the
detector instantaneous field of view (IFOV), minimizing the range gate duration, or by simply operating at night [2]. By
designing and operating the system such that solar background is minimized significantly increases data quality while
increasing the operating range of the sensor and reducing or eliminating the need for noise filtering.
Collection Geometry
The Harris Geiger-mode LiDAR sensor uses a conical Palmer scan pattern produced by a direct drive, hub driven HOE
scanner. HOE scanners have been used in the past [15] but this
particular implementation has several advantages over previous
designs:
Figure 8. Direct drive, hub driven, holographic
optical element scanner
• High power transmit beam is steered by fused silica
transmit prism (1) while low power return signal is collected
by holographic optical element (2) eliminating chance of
damage to holographic element.
• The mechanical axes of the transmit prism (1) and receive
holographic element (2) are coaxial and rotated by the same
direct drive motor.
• Transmit prism diameter determined by out-going laser
beam diameter only (doesn’t scale with receive
aperture).
• Uses smaller diameter bearings compared to perimeter
driven scanners, reducing drag and weight and increasing
reliability.
• Scales to apertures > 300 mm and scan rates of up to 6,000
• Simple mechanical, electrical and control direct
drive
system design with few moving parts.
• Transmit LOS clocking mechanism (3)
• Leverages COTS components, minimizing custom parts
• Low power consumption
•
Low mass, compact, mechanically simple.
The estimated nominal design range of practical scan half angles is 10° to 45° determined by available laser power
which ultimately limits the maximum scan angle for any given altitude and sample density. The current system uses a
scan half angle of 15° which, when flying at 6 kilometers above ground level, the resultant swath width is 3.2 kilometers
wide. Assuming a platform speed of 440 kilometers per hour (240 kts) yields an instantaneous area collection rate of
1,050 square kilometers per hour. The sensor also has automated range gate delay control as well as multi-pulse in flight
discrimination logic.
When flown with 50 percent overlap, the scan pattern provides four looks from four different directions. With a scan half
angle of 15° the sides as well as roofs of structures is sampled, which significantly increases the interpretability of the
point cloud data.
Figure 10. Palmer scan pattern with 50% overlap provides four independent views of the scene
Proc. of SPIE Vol. 9465 946506-6
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
Sensor Performance
The system is capable of imaging a 10% Lambertian target at a spatial sample density of 2 points per square meter at an
altitude of 6 km during the day in an urban aerosol environment with 5 km visibility. When operated at night in a rural
aerosol environment with 23 km visibility the operating altitude is increased to 8 km (Figure 11).
Harris Geiger mode LIDAR
= 90%
p = 10% Lambertian, 150 m Gate Width,
DCR = 2 kHz, PDE = 0.2, Rural Aerosol 23 km Visibility
8
25
0-40 Day Collection Altitude
720
25
.Night Collection Altitude
6
8
Ê
ó5
15
10
S
O
â
4
10
3
10
60
40
samples per square meter
Figure 11. Altitude and area coverage rate versus sample density
Example Imagery
Some examples of imagery from the initial flight tests are shown in Figure 12. Sample density and range precision
support the collection of USGS Quality Level 1 data (8 pts/m2) at an altitude of 6 km and Quality Level 2 (2 pts/m2) at 8
km and above.
111111111111111W
wwww,n,.MT,!plicRWA.11sr7rrn77
"Nlimmwmil1111
Figure 12. Imagery of San Diego CA, single pass from 17 kft
Proc. of SPIE Vol. 9465 946506-7
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms
4.
SUMMARY
Over the past decade and a half, several imaging LiDAR sensors based on Geiger-mode avalanche photodiode
(GmAPD) arrays have been successfully flown providing the technical foundation for a commercial mapping Geigermode LiDAR system. The commercial availability of high PRF, high pulse energy lasers and large format GmAPD
cameras now make the construction of a commercial Geiger-mode LiDAR system economically viable.
REFERENCES
[1]
M. Albota, R. Heinrichs, D. Kocher, D. Fouche, B. Player, M. O’Brien, … R. Carlson, “Three-dimensional
imaging laser radar with photon-counting avalanche photodiode array and microchip laser”, Applied Optics,
41:7671-7678, 2002
[2] F.G. Fouche, “Detection and false-alarm probabilities for laser radars that use Geiger-mode detectors”, Applied
Optics, 42:5388-5398
[3] B. F. Aull, A. H. Loomis, D. J. Young, R. M. Heinrichs, B. J. Felton, P. J. Daniels, D. J. Landers, “Geiger-mode
avalanche photodiodes for three-dimensional imaging”, MIT Lincoln Laboratory Journal 13, 335 – 350 (2002).
[4] R. M. Marino, T. Stephens, R.E. Hatch, J.L. McLaughlin, J.G. Mooney, M. O’Brien, G.S. Rowe, J.S. Adams, L.
Skelly, R.C. Knowlton, S.E. Forman, W.R. Davis, “A compact 3D imaging laser radar system using Geiger-mode
APD arrays: system and measurements”, Proc. of SPIE, vol. 5086, pp. 1-15, 2003.
[5] B. F. Aull, A. H. Loomis, D. J. Young, et al., “Three-dimensional imaging with arrays of Geiger-mode avalanche
photodiodes”, Proc. of SPIE 5353, 105 – 116 (2004).
[6] M. Vaidyanathan, S. Blask, T. Higgins, W. Clifton, D. Davidsohn, R. Carson … R. Heinrichs, “JIGSAW Phase III:
A miniaturized airborne 3-D imaging laser radar with photon-counting sensitivity for foliage penetration”, Proc. of
SPIE 6550 (2007)
[7] R. Cannata, W. Clifton, S. Blask, R.M. Marino, “Obscuration Measurements of Tree Canopy Structure Using a 3D
Ladar Imaging System”, Proc. of SPIE Laser Radar Technology and Applications VIII, vol. 5086, 17 (2003)
[8] Robert Knowlton, “Airborne Ladar Imaging Research Testbed”, MIT Lincoln Laboratory Tech Notes, (2011)
(www.ll.mit.edu/publications/ technotes/ TechNote_ALIRT.pdf)
[9] George Gray, “High Altitude Lidar Operations Experiment (HALOE) – Part 1, System Design and Operation”,
Proc. of Military Sensing Symposium, Active Electro-Optic Systems, paper AH03, (2011)
[10] M. Itzler, M. Entwistle, M. Owens, K. Patel, X. Jiang, K. Slomkowski, S. Rangwala, P.F. Zalud, T. Senko, J.
Tower, J. Ferrar (2010). “Geiger-mode avalanche photodiode focal plane arrays for three-dimensional imaging
LADAR”, Proc. of SPIE, vol. 7808, pp. 78080C-1 – 78080C-14, 2010.
[11] M. A. Itzler, M. Entwistle, M. Owens, K. Patel, X. Jiang, K. Slomkowski, S. Rangwala, P. F. Zalud, T. Senko, J.
Tower, and J. Ferraro, “Comparison of 32 x 128 and 32 x 32 Geiger-mode APD FPAs for single photon 3D
LADAR imaging”, Proc. of SPIE 8033, 80330G (2011).
[12] P. Yuan, R. Sudharsanan, X. Bai ; Joseph Boisvert ; Paul McDonald ; Eduardo Labios ; Bryan Morris ; John P.
Nicholson ; Gary M. Stuart ; Harrison Danny ; Stephen Van Duyne ; Greg Pauls ; Stephen Gaalema, “Geiger-mode
LADAR Cameras”, Proc. of SPIE 8037, 803712 (2011).
[13] X. Jiang, M. A. Itzler, R. Ben-Michael, K. Slomkowski, M. A. Krainak, S. Wu, and X. Sun, “Afterpulsing effects
in free-running InGaAsP single photon avalanche diodes”, IEEE J. Quantum Electron., vol. 44, p. 3 – 11 (2008).
[14] M. A. Itzler, U. Krishnamachari, Q. Chau, X. Jiang, M. Entwistle, M. Owens, K. Slomkowski, “Statistical analysis
of dark count rate in Geiger-mode APD FPAs”, Proc. of SPIE 9250, 925003 (2014).
[15] G. K. Schwemmer, T. D. Wilkerson, D. V. Guerra, “Compact scanning lidar systems using holographic optics” ,
Proc. of SPIE 3504 Optical Remote Sensing for Industry and Environmental Monitoring, 51 (1998).
Proc. of SPIE Vol. 9465 946506-8
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/09/2015 Terms of Use: http://spiedl.org/terms