3D Neural Rendering for Automated Ultrasonic Laser Scans
Abstract
Laser doppler vibrometer (LDV) scans can be used to detect defects and cracks in manufactured parts before they are deployed. This work presents a method for reconstructing the steady-state surface wavefield of a three-dimensional part by combining automated robotic manipulation with neural rendering. A six-degree-of-freedom (6-DOF) robotic manipulator (UFactory xArm6) with an embedded piezoelectric transducer presented an acoustically excited test article to a stationary LDV at a sequence of known orientations, yielding a set of complex-valued 2D scans with poses derived analytically from the arm kinematics rather than from image-based pose estimation. These scans were used to train a novel complex-nerfacto model, a modified Neural Radiance Field (NeRF) pipeline that replaces the standard three-channel RGB output head with a two-channel output. Complex-nerfacto represents the real and imaginary velocity components of a wavefield, preserving the complex measurement without quantization. Reconstructions of a welded aluminum cylinder and a welded T-bracket were evaluated against measured scans using Z-normalized cross-correlation (ZNCC). The complex-valued pipeline achieved a mean ZNCC of 0.60 at held-out viewing angles, compared with 0.28 for a baseline RGB nerfacto model trained on rendered images of the same data. The resulting surface wavefield reconstructions supported defect detection and characterization on complex components. We identify geometry reconstruction from sparse LDV scans as the remaining obstacle to practical deployment.
Related
- Laser ultrasonic investigation of chromium coating impact on elastic guided waves in zirconium tubes
- Automated CR-39 pit detection for energy and angular characterization of laser-driven ion beams
- Rapidity-Coupled Spin Dynamics in Pulsed Laser Fields from Physics-Informed Neural Networks
- Automated Dislocation Detection in Electron Channelling Contrast Imaging: A Comparative Study of Rule-Based, Neural Network, and Deep Learning Approaches
- Automated AFGL quantum number assignment for CO2 isotopologues using a graph neural network
BibTeX
@misc{pierson2026neural,
title = {{3D Neural Rendering for Automated Ultrasonic Laser Scans}},
author = {Thomas Pierson and Matthew Mitchell and Kira Perna and Anand Iyer and Eric Flynn and Adam Wachtor},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.00527},
eprint = {2611.00527},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.00527},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - 3D Neural Rendering for Automated Ultrasonic Laser Scans
AU - Thomas Pierson
AU - Matthew Mitchell
AU - Kira Perna
AU - Anand Iyer
AU - Eric Flynn
AU - Adam Wachtor
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.00527
UR - https://arxiv.org/abs/2611.00527
AB - Laser doppler vibrometer (LDV) scans can be used to detect defects and cracks in manufactured parts before they are deployed. This work presents a method for reconstructing the steady-state surface wavefield of a three-dimensional part by combining automated robotic manipulation with neural rendering. A six-degree-of-freedom (6-DOF) robotic manipulator (UFactory xArm6) with an embedded piezoelectric transducer presented an acoustically excited test article to a stationary LDV at a sequence of known orientations, yielding a set of complex-valued 2D scans with poses derived analytically from the arm kinematics rather than from image-based pose estimation. These scans were used to train a novel complex-nerfacto model, a modified Neural Radiance Field (NeRF) pipeline that replaces the standard three-channel RGB output head with a two-channel output. Complex-nerfacto represents the real and imaginary velocity components of a wavefield, preserving the complex measurement without quantization. Reconstructions of a welded aluminum cylinder and a welded T-bracket were evaluated against measured scans using Z-normalized cross-correlation (ZNCC). The complex-valued pipeline achieved a mean ZNCC of 0.60 at held-out viewing angles, compared with 0.28 for a baseline RGB nerfacto model trained on rendered images of the same data. The resulting surface wavefield reconstructions supported defect detection and characterization on complex components. We identify geometry reconstruction from sparse LDV scans as the remaining obstacle to practical deployment.
ER -