Multispectral precision distance metrology and material probing

Recent advances in mode-locked femtosecond (fs) lasers and spectral broadening techniques enable new approaches to multi- or hyperspectral precision metrology. In a Swiss National Science Foundation (SNSF) co-funded project, we demonstrate the application of supercontinuum (SC) sources derived from mode-locked fs lasers for multiwavelength electro-optical distance measurement (EDM) and remote surface material probing.

The SC spectrum covers hundreds of nm as compared to the few tens of nm of the seed mode-locked laser (MLL), thus allowing spectral flexibility for distance measurements over several wavelengths from a single laser source. The repetition-rate of the MLL is locked to an Rb frequency standard, which we also use to establish a common time base for the down-conversion and digitizing electronics. We experimentally investigate the use of intermode beats generated upon direct photodetection of an optically filtered spectrum of the SC for distance measurements. This avoids the need for any optical or electrical modulator and reduces instrumental errors due to crosstalk. Power and delay differences between the spectral ranges can in turn be used to derive information about the propagation media and the remote target.

Enlarged view: Experimental Setup — Picture of the optical table in our laboratory showing the developed experimental platforms.

This investigation may potentially lead to an instrumental basis for future generations of high-precision EDM and LiDAR instruments, which would not only benefit conventional surveying and 3D reality capture but also enable augmented LiiDAR-based material sensing beyond 3D in various application contexts such as automated construction, cultural heritage preservation, circular construction, mining, forest management, and smart agriculture.

Our research has so far demonstrated:

i) Refractivity-corrected distance measurements using two filtered spectral bands of a SC, to achieve highly precise distance measurements over long distances on a cooperative target. (Ref: [5])

Enlarged view: Refractivity correction — Time series showing (a) pressure variations of P monitored along the optical path (b) corresponding distance deviations measured using the reference interferometer ΔLHeNe, (c) simultaneously acquired optical path length deviations on the two wavelengths Δℓλ1 and Δℓλ2 (which is shifted by −0.1 mm for clarity), two-color refractivity compensated measurements ΔD, and its 21-point moving-averaged result ΔD. [5]

ii) Hyperspectral LiDAR for precision laser scanning, indicating the potential to enable finer point cloud segmentation, spectrum-based material classification, and vegetation health monitoring. (Ref: [4, 7]) The precise range information additionally enables improved radiometric correction of backscatter intensity (Ref: [4, 11])

Enlarged view: Hyperspectral Lidar — 3D point cloud of the wooden target showing (left) geometrical profile, and (right) material classification using spectral information. Z-axis represents the range information for the target placed in the x-y plane [7]

Enlarged view: Vegetation health monitoring — Point-wise mapping of the (a) NDVI and (b) RVI values calculated using the acquired hyperspectral information, after correcting the angle of incidence-related bias in the acquired intensities [4]

iii) Multimodal multispectral (MM) LiDAR sensing and a specialized feature selection method (Ref: [3, 8])

Enlarged view: MM Lidar — Multimodal multispectral (MM) features of five material specimens in the form of reflectance (R), distance (d), and polarization (DoLP) spectra. MM LiDAR combines LiDAR, remote spectroscopy, and multimodal sensing technologies into a single instrument, thus providing enriched material information and further augmenting remote 3D digitization and environmental perception. The remote spectroscopy applied herein covers 33 spectral channels centered between 580 nm and 900 nm. Multimodal sensing applied herein consists of reflectance, distance, and degree of linear polarization modalities, which describe light absorption, penetration, and depolarization during light-matter interaction.

Enlarged view: Feature selection — The feature selection method for MM LiDAR sensing and its performance. The proposed multiclass group feature selection algorithm based on an all-in-one support vector machine (MGSVM FS) tackles the challenges of feature selection for structural data and multiclass classification problems in MM LiDAR sensing. Left: Illustration of absolute coefficient values output from MGSVM FS. The coefficients with non-zero values (red boxes) recommend the spectral channels important in the multiclass classification task using multiple optical modalities. Right: Maximally achievable classification performance (mean F1-score) and the required number of spectral channels for all modality combinations. The involvement of more modalities requires fewer spectral channels to achieve the best classification performance. [3]

iv) Augmented classification of material and surface roughness using new polarization-resolved reflectance spectra. (Ref: [6])

Enlarged view: Polarized reflectance — The standard reflectance (R), unpolarized reflectance (Runpol) and linearly polarized reflectance (Rpol) spectra of ten material specimens. The ten specimens consist of five materials and two levels of surface roughness. The unpolarized reflectance spectrum outperforms the standard reflectance spectrum in material classification. The linearly polarized reflectance spectrum outperforms the standard reflectance spectrum in roughness classification. [6]

List of publications:

Doctoral Theses:

[1] P. Ray, Hyperspectral long-distance metrology using a femtosecond laser supercontinuum, Doctoral Thesis, ETH Zurich (2024), doi:10.3929/ethz-b-000667524

[2] Y. Han, Material probing using multimodal multispectral LiDAR based on a femtosecond laser supercontinuum, Doctoral Thesis, ETH Zurich (2024), doi: 10.3929/ethz-b-000666007

Journal papers:

[3] Y. Han, D. Salido-Monzú, J.A. Butt, S. Schweizer, and A. Wieser, A feature selection method for multimodal multispectral LiDAR sensing, ISPRS Journal of Photogrammetry and Remote Sensing, Vol. 212, pp. 42-57 (2024), doi: 10.1016/j.isprsjprs.2024.04.022

[4] P. Ray, T. Medic, D. Salido-Monzú, and A. Wieser, High-precision hyperspectral laser scanning for improved radiometric correction of backscatter intensity, Optical Engineering, accepted, (2024)

[5] P. Ray, D. Salido-Monzú, R. Presl, J.A. Butt, and A. Wieser, Refractivity corrected distance measurement using the intermode beats derived from a supercontinuum, Optics Express, Vol. 32, Issue 7, pp. 12667-12681 (2024), doi:10.1364/OE.514997

[6] Y. Han, D. Salido-Monzú, and A. Wieser, Classification of material and surface roughness using polarimetric multispectral LiDAR, Optical Engineering, Vol. 62, Issue 11, 114104 (2023), doi: 10.1117/1.OE.62.11.114104

[7] P. Ray, D. Salido-Monzú, S.L. Camenzind, and A. Wieser, Supercontinuum-based hyperspectral LiDAR for precision laser scanning, Optics Express, Vol.31, Issue 20, pp. 33486-33499 (2023), doi:10.1364/OE.498576

[8] Y. Han, D. Salido-Monzú, and A. Wieser, Comb-based multispectral LiDAR providing reflectance and distance spectra, Optics Express, Vol. 30, Issue 23, pp. 42362-42375 (2022), doi: 10.1364/OE.473466

[9] P. Ray, D. Salido-Monzú, and A. Wieser, High-precision intermode beating electro-optic
distance measurement for mitigation of atmospheric delays, Journal of Applied Geodesy,
Vol. 17, Issue 2, pp. 93-101 (2023), doi:10.1515/jag-2022-0039

Conference papers:

[10] Y. Han, D. Salido-Monzú, and A. Wieser, "Classification of material and surface roughness using polarimetric multispectral LiDAR," in Multimodal Sensing and Artificial Intelligence: Technologies and Applications III, 12621, 12621D, SPIE (2023).

[11] P. Ray, D. Salido-Monzú, A. Wieser, and T. Medic, Supercontinuum-based hyperspectral laser scanning: towards enhanced 3d surface reconstruction and its benefits for remote sensing, Multimodal Sensing and Artificial Intelligence: Technologies and Applications III, Vol. 12621 (2023)

[12] Y. Han, D. Salido-Monzú, J.A. Butt, and A. Wieser, Polarimetric femtosecond-laser LiDAR for multispectral material probing, Optics and Photonics for Advanced Dimensional Metrology II, Vol 12137 (2022)

[13] Y. Han, D. Salido-Monzú, and A. Wieser, Delay-Augmented Spectrometry for Target Classification Using a Frequency-Comb LiDAR, CLEO: Science and Innovations, SF2F. 5 (2022)

[14] P. Ray, D. Salido-Monzú, and A. Wieser, High-precision intermode beating EDM for mitigation of atmospheric delays, 5th Joint International Symposium on Deformation Monitoring -JISDM (2022)

Contact

Prof. Dr. Andreas Wieser

Full Professor at the Department of Civil, Environmental and Geomatic Engineering

Geosensorik und Ingenieurgeodäsie
Stefano-Franscini-Platz 5
8093 Zürich
Switzerland