Effect of incidence angle on laser scanner intensity and surface data

OSAPublished by

To be published in Applied Optics:

Title: Effect of incidence angle on laser scanner intensity and surface data

Authors: Antero Kukko, Sanna Kaasalainen, and Paula Litkey

Accepted: 4 January 2008

Posted: 18 January 2008

Doc. ID: 86812

OSAPublished by

1

Effect of incidence angle on laser scanner intensity and

surface data

Antero Kukko,1 Sanna Kaasalainen,1 and Paula Litkey1

1 Department of Remote Sensing and Photogrammetry, Finnish Geodetic Institute, PO Box 15

FIN-02431 Masala, Finland

OSAPublished by

2

We present a comprehensive experimental set of data on the dependence of the laser

intensity on the angle of incidence to the target surface. The measurements have been

performed in laboratory for samples with a Nd:YAG laser and terrestrial laser scanner

(TLS). The brightness scale data were also compared with data acquired by airborne laser

scanning (ALS). The incidence angle effect is evident for all the targets. The effect is

significant for incidence angles greater than 20°, and stronger for bright targets. However,

effects due to some of the other surface properties, such as roughness, were also detected.

We also found a set of gravel samples, for which the incidence angle effect was minor

even up to 40°. The data provide important reference for the interpretation and

applications, e.g., full-waveform data processing of a laser scanner and ALS intensity

calibration.

© 2007 Optical Society of America

OCIS codes: 280.3640, 280.1350, 290.5820.

1. Introduction

Lidar systems have become a well-established tool for providing accurate 3D-information on

remote sensing targets. There is abundant knowledge on system uncertainties and their effect on

the topographic models, but there are little studies on the role of the optical properties of the

targets and the laser light interaction with the surface media. A number of studies on the light

interaction with surfaces and diffusive media are available [1,2], and the effect of the optical

properties of different targets on their backscatter [3,4,5,6]. Simulations on the laser beam

interaction with the target surface have provided some important information on the properties

OSAPublished by

3

with strongest effects on the laser interaction and the interpretation of the backscattered pulse [7,

8], but there is a growing need for experimental data on the factors affecting the laser echo, such

as the surface composition and structure, and especially the orientation of the target. Systematic

experimental data on the effect of the incidence (i.e., scanning) angle on the intensity of typical

laser scanner land targets provide important reference for the interpretation and applications, e.g.,

simulations of full-waveform laser scanner data.

The observing geometry in all the current lidar systems is basically the same: the

returning pulse is recorded practically at backscatter, i.e. the light paths of the source and

detector coincide. This geometry is a source of some important scattering effects that are studied

in many fields of optics and physics [1,3,4,5,6] and references therein, but since the observing

geometry remains the same in all lidar applications, the effects from the change of geometry are

mostly avoided.

The angle of incidence to the target plays an important role in both the 3D-model and the

backscattered intensity recorded by the lidar sensor. The effect of incidence angle on backscatter

has been addressed computationally and experimentally in numerous works in photonics and

optics [e.g. 1,2,9,10,11]. These studies have mostly concentrated on specific case studies, e.g. the

polarized backscatter from diffusive media, or fiber on a reflective surface [1,2,9,10], but the

applications of these results to remote sensing of natural surfaces are limited. No systematic

experimental data on the effect of incidence angle on different materials, especially the typical

remote sensing targets, are available [1]. Such data would also be important for the investigation

of lidar-based reflectance measurement, which has recently become a topic of scientific interest

[12,13,14,15,16]. One of the few attempts to correct the incidence angle effect on lidar intensity

is based on simple cosine approach where a Lambertian surface is assumed [17,18].

OSAPublished by

4

We present a systematic set of data on the dependence of the laser intensity on the angle

of incidence to the target surface. To our knowledge this is the first comprehensive study on

these effects in the context of scanning lidars. Data from independent measurement systems

(laboratory and terrestrial laser scanner) are presented and compared, and the implications of the

results on lidar-derived 3D-models (such as the possible empirical corrections and applications)

are discussed.

2. Methods

2.1 The laboratory experiment

The laboratory laser instrument has been constructed to operate in the similar

illumination/observation geometry as in laser scanning (i.e., exact backscatter where the source

and detector light paths coincide). The instrument (sketched in Fig. 1) comprises a 1064 nm

Nd:YAG laser (wavelength similar to most airborne scanners), and 16-bit monochrome CCD-

camera (Sbig ST-7), which is a commonly used detector in laboratory (laser) measurements in,

e.g., optical physics [5]. We averaged five 3-second images for each target. To illuminate a

larger area of the sample than the 3mm laser spot (and hence reduce the deviation in data), the

samples were placed on a rotator allowing the laser spot to move around the sample surface

during the image exposures. The backscattered laser intensities were measured from the CCD

images by means of standard photometric techniques. More details on the experiment are found

in [12,13]. The incidence angles were changed using a precision goniometer to tilt the sample

from normal (0°) incidence (see Figs. 1 and 2).

To demonstrate the wavelength effects on the incidence angle dependence in general, we

also carried out a hyperspectral measurement for some of the targets. The measurement was

OSAPublished by

5

made similarly to the 1064 nm experiment (i.e. at the exact backscatter angle) by replacing the

Nd:YAG with a supercontinuum (Koheras SuperK Red) laser and installing a spectrograph

(Specim Imspector) in front of the CCD camera. The spectra were measured at the wavelength

range of 600-900 nm. The instrument was similar to that presented in [19] with a better

alignment and thus improved accuracy.

2.2 Terrestrial and airborne laser scanner measurements

The validation of the results of the laboratory laser measurements was carried out with the

FARO LS 880HE80 (FARO LS) terrestrial laser scanner operating at 785nm. The scanner uses

phase angle technique for the distance measurement with the accuracy of 3-5 mm and 360°×320°

field of view. The detector of the FARO LS is not optimized for intensity measurement: there are

modifications in the detector that affect the intensity, e.g., a brightness reducer for near distances

(<10 m) and a logarithmic amplifier for small reflectances. These all required extensive and

systematic distance and reflectance calibrations, which were carried out in the laboratory using a

calibrated grayscale and a calibrated 4-step Spectralon reflectance panel [12,13]. The incidence

angle measurements were carried out at 1-meter target distance placing the sample on the

goniometer for the incidence angle variation (Fig. 2).

Three of the brightness scale targets with 10%, 20%, 50% nominal reflectances were also

measured at the Espoonlahti ALS flight campaign, conducted Aug 31st, 2006 using the TopEye

MKII 1064 nm laser scanner. The flying altitude was 300 meters and the test area consisted of

the Espoonlahti boat harbor and beach. 20-100 hits (i.e., returned pulses) were recorded from

each sample.

2.3 The samples

OSAPublished by

6

We investigated a set of targets used in airborne laser scanner intensity calibration, and a set of

natural and artificial (such as brick material) samples representing typical airborne laser scanner

land targets.

The brightness calibration targets (tarps) measured in this study have been used in

radiometric calibration of aerial cameras and airborne laser scanners. The tarps are made of

polyester fiber with PVC coating and span a brightness grayscale from 5% to 70% of reflectance

[15]. We measured a set of four targets with calibrated reflectances of 8%, 26%, 50%, and 70%

[12]. All measurements are relative to the Spectralon 99% reference target at 0° incidence. The

incidence angle effect of the Specralon target was also measured.

We have also studied the possibility of using standard industrial gravels in airborne laser

scanner intensity calibration [12,13]. A set of gravels with distinctly different reflectances, small

deviation and difference between laboratory and laser scanner measurement, and little variation

with the measurement geometry would be ideal for the purpose of intensity calibration.

Preliminary tests have shown that gravels with uniform color range with respect to the

dimensions of the measurement (i.e., the laser spot/footprint size) are most suitable for laboratory

reference measurements. As the laboratory brightness calibration is essential for the use of these

gravels as intensity standards in airborne laser scanning campaigns, it is important that the results

are consistent even for the small-scale (laboratory) measurements. The footprint of an airborne

laser scanner might allow some larger scale variations in the target structure and brightness, but

reproducing such results in laboratory (where a collimated laser source is needed to reproduce

the backscatter measurement geometry of laser scanners e.g. [12,13,19]) has proven a challenge.

The sands and gravels chosen for this study are commonly used in sandblasting and

construction. We measured sandblasting sand of two grain sizes, black gabro gravel, crushed

OSAPublished by

7

redbrick and LECA® (Light Expanded Clay Aggregate), which consists of small, lightweight,

bloated particles of burnt clay. These gravels were also chosen because of their easy availability

in standard hardware stores, which would be practical for the calibration of airborne laser data

taken at different campaigns in varying locations. The preliminary test measurements for a large

set of gravels showed that it is difficult to find a gravel of high reflectance suitable for

calibration. Therefore, to extend the brightness scale towards higher reflectances, we included a

sample of polystyrene into the investigation. As the gravel samples started spilling at high tilt

angles, the measurements over 40° of incidence were not practical. To demonstrate the incidence

angle effects for natural samples, we also present a spectral result for a linden (Tilia Cordata)

leaf.

3. Results and Discussion

The incidence angle and brightness functions for the brightness tarps and gravels, measured with

the laboratory instrument and the FARO LS, are presented in Figs. 3 and 4. The change in

brightness with the incidence angle is practically negligible for most targets at the incidence

angle range up to 20º (and even up to 30º for targets with reflectance < 50%). For the brightest

targets (Spectralon 99% and the polystyrene target, see Fig. 5) the decrease in brightness is

significant even at small angles of incidence. The incidence angle effect seems to be stronger for

the targets of high reflectance, but the effect of other parameters (such as the surface roughness

scale compared to the laser spot size) should be further investigated for quantitative results. The

effect of surface roughness on brightness variation can be reviewed qualitatively in the 3D

structure plots in Fig 3. It seems that other features than surface roughness play stronger role in

the incidence angle effect. For example, the crushed redbrick sample shows strong variation in

overall intensity, which is partly due to the large grain size (1-2 cm) compared to the laser spot

OSAPublished by

8

size (5 and 3 mm for Nd:YAG and FARO, respectively), but the degrease in intensity is not

stronger than that of other gravels with smaller grain size (see Fig. 3).

The decrease in brightness gets more significant when the angle of incidence increases

from 20º up to 70º. For comparison, [2] found a decrease in the backscatter peak as the angle of

incidence increases. As our measurements were made in the backscatter direction, this could at

least partially explain the decrease in the intensity towards 70º angle of incidence. However, the

role of specular reflection at normal (0º) incidence should be further investigated in this case.

The SuperK hyperspectral measurements enabled us to investigate the possible

wavelength effects. Figs. 5 shows the spectra of polystyrene at 0º and 30º. The intensity level has

dropped, but the shapes of the spectra are the same. This can also be seen in the 3D spectral

images (Fig. 6) for linden, crushed redbrick, and the sandblasting sand: there is little variation in

the shape of the spectrum, even though the intensity level changes as a function of the incidence

angle.

The results (i.e. the overall trends in intensity) obtained in laboratory measurements are

in agreement with those measured with the FARO terrestrial laser scanner in spite of the

wavelength difference between the instruments. This means that the brightness effects related to

laser scanning can be simulated in the laboratory and the results can be used in the interpretation

of ALS brightness and surface data and their calibration. Fig. 7 shows a further comparison of

different measurements for same targets. The wavelength effects are evident (i.e. there are offsets

in the brightness levels between different wavelengths), but the overall brightness trends with

incidence angle are reproduced.

The comparison of the TopEye ALS and FARO TLS data are presented in Table 1. The

reflectances are expressed in relation to the 70% brightness target with the same angle of

OSAPublished by

9

incidence as the samples. These results also point out that the relative results are fairly well

reproduced at small (up to 20º) angles of incidence. There are differences between ALS and TLS

intensity levels, some of which may be explained by the difference in wavelength, but much

more data from ALS campaigns are needed to make a more accurate comparison and to establish

the accuracy limits of ALS intensity measurement [12,13,19]. Other important future issues

include, e.g., the role of multiple and specular reflections. Because of the grainy nature of most

of the samples studied here, the specular reflections seem to occur in all directions, not only in

the normal (0º) incidence.

4. Conclusion

We have investigated the effect of incidence angle on target brightness. Thus far there has been

little of such information available, especially for laser scanning applications, where the

incidence angle plays a crucial role on the intensity calibration. This study gives an overall view

on the incidence angle effects on such targets from a practical point of view. We have also

carried out a hyperspectral study of the incidence angle effects and demonstrated that the

incidence angle effects are consistent between different wavelengths, although there are changes

in the intensity levels of the targets.

We found that the decrease in target brightness is significant mostly at incidence angles

greater than 30°. These angles have little relevance in the possible ALS based radiometric

calibration measurement, but must be taken into account in the interpretation of airborne or TLS

data where steep slopes and vertical surfaces are present in the target area. It also seems that the

target brightness plays some role in the brightness decrease, especially towards large (up to 70°)

angles of incidence.

OSAPublished by

10

We also present a set of gravel samples, for which the decrease in brightness with the

incidence angle was negligible up to ~40º. This makes them ideal targets for airborne laser

scanner brightness calibration, since the typical incidence (i.e., scanning) angle variation is

usually within this range. The laboratory measurements show a gradual increase in the brightness

for these targets, which provide us a field brightness scale from 0.1-0.5. Adding a sample of

polystyrene to the set of gravels, the scale can be extended up to 0.8.

The incidence angle data derived in this study are to be applied in the data analysis for

scanning lidar sensors. Incidence angle affects strongly the achievable range accuracy, as the

shape of the detected echo (of the transmitted pulse) changes with the angle of incidence, unless

proper signal processing algorithms are used as discussed in [20,21]. It is also expected that the

laser scanner data could be further analyzed if such sources of variation are properly understood.

Incidence angle data are furthermore essential in establishing a simulation model for laser

scanning as proposed by [7]. As the systematic experimental data on incidence angle effects are

still in sparse supply, our results apply directly to the optical studies of directional light

scattering.

Acknowledgements

The authors would like to thank Altti Akujärvi for laboratory assistance, and Dr. Hannu Hyyppä

at Helsinki University of Technology for support, photos, and the bulk of music to make the

measurements more comfortable.

References

1. J. Ellis, P. Caillard, and A. Dogariu, “Off-diagonal Mueller matrix elements in backscattering

from highly diffusive media,” J. Opt. Soc. Am. A 19, 43-48 (2002).

OSAPublished by

11

2. V. A. Ruiz-Cortés, and J. C. Dainty, “Experimental light-scattering measurements for large-

scale composite random rough surfaces,” J. Opt. Soc. Am. A 19, 2043-2052 (2002).

3. M. P. van Albada, and A. Lagendijk, “Observation of weak localization of light in a random

medium,” Phys. Rev. Lett. 55, 2692-2695 (1985).

4. D. S. Wiersma, M. P. van Albada, B. A. van Tiggelen, and A. Lagendijk, “Experimental

evidence for recurrent multiple scattering events of light in disordered media,” Phys. Rev. Lett.

74, 4193-4196 (1995).

5. G. D. Yoon, N. G. Roy, and R. C. Straight, “Coherent backscattering in biological media:

Measurement and estimation of optical properties,” Appl. Opt. 32, 580-585 (1993).

6. S. Kaasalainen, J. Peltoniemi, J. Näränen, J. Suomalainen, F. Stenman, and M. Kaasalainen,

”Small-angle goniometry for backscattering measurements in the broadband spectrum,” Appl.

Opt. 44, 1485-1490 (2005).

7. A. Kukko and J. Hyyppä, ”Laser scanner simulator for system analysis and algorithm

development: a case with forest measurements,” Proceedings of the ISPRS Workshop on Laser

Scanning 2007 and SilviLaser 2007 Espoo, September 12-14, 2007, Finland. International

Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences 36, Part 3/W52,

234-240 (2007).

8. B. Jutzi, J. Neulist, and U. Stilla, “High-Resolution waveform acquisition and analysis for

pulsed laser,” in High-resolution earth imaging for geospatial information, C. Heipke, K.

Jacobsen and M. Gerke, eds. International Archives of Photogrammetry, Remote Sensing and

Spatial Information Sciences 36, Part 1/W3, (2005).

OSAPublished by

12

9. G. Videen, W. S. Bickel, V. J. Iafelice, and D. Abromson. Experimental light-scattering

Mueller matrix for a fiber on a reflecting optical surface as a function of incident angle. J. Opt.

Soc. Am. A 9, 312-315 (1992).

10. J. P. Landry, J. Gray, M. K. O'Toole, and X. D. Zhu, "Incidence-angle dependence of optical

reflectivity difference from an ultrathin film on solid surface," Opt. Lett. 31, 531-533 (2006).

11. B. Jutzi, B. Eberle, and U. Stilla, “Estimation and measurement of backscattered signals from

pulsed laser radar,” Proc. SPIE 4885, 256-267 (2003).

12. S. Kaasalainen, A. Kukko, T. Lindroos, P. Litkey, H. Kaartinen, J. Hyyppä, and E. Ahokas,

”Brightness measurements and calibration with airborne and terrestrial laser scanners,” IEEE

Trans. Geosci. Remote Sensing, (in press).

13. S. Kaasalainen, J. Hyyppä, P. Litkey, H. Hyyppä, E. Ahokas, A. Kukko, and H. Kaartinen,

”Radiometric calibration of ALS intensity,” Proceedings of the ISPRS Workshop on Laser

Scanning 2007 and SilviLaser 2007 Espoo, September 12-14, 2007, Finland. International

Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences 36, Part 3/W52,

201-205 (2007).

14. S. Kaasalainen, E. Ahokas, J. Hyyppä, and J. Suomalainen, ”Study of surface brightness

from backscattered laser intensity: calibration of laser data,” IEEE Geoscience and Remote

Sensing Letters 2, 255-259 (2005).

15. E. Ahokas, S. Kaasalainen, J. Hyyppä, and Suomalainen, J., ”Calibration of the Optech

ALTM 3100 laser scanner intensity data using brightness targets,” ISPRS Commission I

Symposium, July 3-6, 2006, Marne-la-Vallee, France, International Archives of

Photogrammetry, Remote Sensing and Spatial Information Sciences 36, Part 1, (2006).

OSAPublished by

13

16. D. S. Boyd and R. A. Hill, “Validation of airborne lidar intensity values from a forested

landscape using hymap data: preliminary analyses,” Proceedings of the ISPRS Workshop on

Laser Scanning 2007 and SilviLaser 2007 Espoo, September 12-14, 2007, Finland. International

Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences 36, Part 3/W52,

71-76 (2007).

17. F. Coren and P. Sterzai, “Radiometric correction in laser scanning,” International Journal of

Remote Sensing 27, 3097-3104 (2006).

18. B. Höfle and N. Pfeifer, “Correction of laser scanning intensity data: Data and model-driven

approaches,” ISPRS Journal of Photogrammetry and Remote Sensing (in Press).

19. S. Kaasalainen, T. Lindroos, and J. Hyyppä, ”Toward hyperspectral lidar - Measurement of

spectral backscatter intensity with a supercontinuum laser source,” IEEE Geoscience and Remote

Sensing Letters 4, 211- 215 (2007).

20. P. Palojärvi, “Integrated electronic and optoelectronic circuits and devices for pulsed time-of-

flight laser rangefinding”. PhD Thesis, Department of Electrical and Information Engineering

and Infotech Oulu, University of Oulu, 54 p (2003).

21. K.-H. Thiel and A. Wehr, “Performance capabilities of laser scanners – an overview and

measurement principle analysis,” International Archives of Photogrammetry, Remote Sensing

and Spatial Information Sciences 36, Part 8/W2, 14-18 (2004).

OSAPublished by

14

List of figure captions

Fig. 1.The laboratory laser (Nd:YAG) measurement arrangement. The sample surface was tilted

using a goniometer (see also Fig. 2) to change the incidence angle.

Fig. 2. (a) The horizontal laboratory arrangement for the FARO measurement of the incidence

angle. (b) The goniometer used to deflect the target to gain the incidence angle range 0°-70°.

Image courtesy by Hannu Hyyppä.

Fig. 3. Relative reflectances plotted as a function of incidence angle for the (a) Gabro, and (b)

LECA gravels, (c) crushed redbrick, and (d), (e) sandblasting sand of two different grain sizes

(0.1-0.6 mm and 0.5-1.2 mm). The left panels were measured with the 1064 nm laboratory

instrument and the right panels with the FARO terrestrial laser scanner. In the middle panels, the

3D structures of the samples were measured with Konica-Minolta VI-9i laser digitizer. The

sample images are in the same scale.

Fig. 4. Relative reflectances plotted as a function of incidence angle for the 99% Spectralon

reflectance target (a) and the PVC brightness calibration tarps 70%, 50%, 26%, and 8% (b-e,

respectively). The left panels were measured with the 1064 nm laboratory instrument and the

right panels with the FARO terrestrial laser scanner.

Fig. 5. The spectra (a and b) and relative reflectances as a function of incidence angle (c, d, e) for the

polystyrene sample. The spectra are plotted at (a) 0º and (b) 30º angles of incidence. The decrease in

OSAPublished by

15

brightness is obvious while the shape of the spectrum does not change significantly. The incidence

angle plots are presented at 633 nm (c), 785 nm (d), and 900 nm (e).

Fig. 6. The variation of brightness with incidence angle plotted over 600-900 nm for linden (a),

crushed redbrick (b), and the 0.1-0.6 mm sandblasting sand (c).

Fig. 7. Comparison of the results measured with different sensors and light sources for the

sandblasting sand (0.1-0.6 mm) (a), crushed redbrick (b), and the 50% test target (c). From the

left: FARO (785 nm), SuperK (at 785 nm), Nd:YAG (1064 nm), and SuperK (at 900 nm). There

are brightness level variations (especially for the redbrick, whose spectrum shows a strong

increase in brightness towards the near-infrared end of the spectrum), but the incidence angle

effects are consistent between different measurements.

OSAPublished by

16

Fig. 1.The laboratory laser (Nd:YAG) measurement arrangement. The sample surface was tilted

using a goniometer (see also Fig. 2) to change the incidence angle.

OSAPublished by

17

Fig. 2. (a) The horizontal laboratory arrangement for the FARO measurement of the incidence

angle. (b) The goniometer used to deflect the target to gain the incidence angle range 0°-70°.

Image courtesy by Hannu Hyyppä.

OSAPublished by

18

Fig. 3. Relative reflectances plotted as a function of incidence angle for the (a) Gabro, and (b)

LECA gravels, (c) crushed redbrick, and (d), (e) sandblasting sand of two different grain sizes

(0.1-0.6 mm and 0.5-1.2 mm). The left panels were measured with the 1064 nm laboratory

instrument and the right panels with the FARO terrestrial laser scanner. In the middle panels, the

3D structures of the samples were measured with Konica-Minolta VI-9i laser digitizer. The

sample images are in the same scale.

OSAPublished by

19

Fig. 4. Relative reflectances plotted as a function of incidence angle for the 99% Spectralon

reflectance target (a) and the PVC brightness calibration tarps 70%, 50%, 26%, and 8% (b-e,

respectively). The left panels were measured with the 1064 nm laboratory instrument and the

right panels with the FARO terrestrial laser scanner.

OSAPublished by

20

Fig. 5. The spectra (a and b) and relative reflectances as a function of incidence angle (c, d, e) for the

polystyrene sample. The spectra are plotted at (a) 0º and (b) 30º angles of incidence. The decrease in

brightness is obvious while the shape of the spectrum does not change significantly. The incidence

angle plots are presented at 633 nm (c), 785 nm (d), and 900 nm (e).

OSAPublished by

21

Fig. 6. The variation of brightness with incidence angle plotted over 600-900 nm for linden (a),

crushed redbrick (b), and the 0.1-0.6 mm sandblasting sand (c).

OSAPublished by

22

Fig. 7. Comparison of the results measured with different sensors and light sources for the

sandblasting sand (0.1-0.6 mm) (a), crushed redbrick (b), and the 50% test target (c). From the

left: FARO (785 nm), SuperK (at 785 nm), Nd:YAG (1064 nm), and SuperK (at 900 nm). There

are brightness level variations (especially for the redbrick, whose spectrum shows a strong

increase in brightness towards the near-infrared end of the spectrum), but the incidence angle

effects are consistent between different measurements.

OSAPublished by

23

List of Tables

Table 1. Comparison of ALS (TopEye, 1064 nm) and TLS (FARO, 785 nm) relative reflectances

for 50%, 20 %, and 10% test targets. The results are relative to the 70% test target. For the ALS

data the incidence angles were 12°, 18° and 20°, and the TLS data was chosen accordingly.

OSAPublished by

24

Table 1. Comparison of ALS (TopEye, 1064 nm) and TLS (FARO, 785 nm) relative reflectances

for 50%, 20 %, and 10% test targets. The results are relative to the 70% test target. For the ALS

data the incidence angles were 12°, 18° and 20°, and the TLS data was chosen accordingly.

Tarp TopEye 12° TopEye 18° TopEye 20° FARO 12° FARO 18° FARO 20°

50% 0.68 0.68 0.72 0.79 0.79 0.78

20% 0.36 0.37 0.38 0.24 0.24 0.23

10% 0.12 0.12 - 0.12 0.12 0.12