Kinematic target surface sensing based on improved deep optical flow tracking
An improved deep optical flow network tracks kinematic target surfaces frame-to-frame in structured light sequences, enabling accurate surface sensing of moving mechanical components without requiring hardware synchronisation or target marking.
Nov 1, 2023
Deep learning-enabled 3D multimodal fusion of cone-beam CT and intraoral mesh scans for clinically applicable tooth-bone reconstruction
A deep learning framework fuses cone-beam CT volumetric data with intraoral optical mesh scans to produce clinically accurate 3D tooth-bone reconstructions, enabling precise pre-surgical planning without manual segmentation.
Sep 1, 2023
Towards uniform point distribution in feature-preserving point cloud filtering
A point cloud filtering method simultaneously preserves sharp geometric features and promotes spatially uniform point distribution, resolving the typical trade-off where feature-preserving filters cluster points near sharp edges.
Jan 1, 2023
Random screening-based feature aggregation for point cloud denoising
Random screening selects informative subsets of neighbouring points for feature aggregation during point cloud denoising, reducing sensitivity to outliers and improving denoised geometry quality over fixed-neighbourhood aggregation schemes.
Jan 1, 2023
3D reconstruction of the moving object with double-shooting based on phase shifting profilometry
This work focuses on 3d reconstruction of the moving object with double-shooting based on phase shifting profilometry, contributing to Wei Pan's research thread in 3D vision, optical metrology, geometry processing, and industrial inspection.
Jan 1, 2023
Automated reconstruction of multiple objects with individual movement based on PSP
An automated pipeline reconstructs 3D shapes of multiple independently moving objects simultaneously using phase-shifting profilometry, tracking each object separately to enable motion compensation and per-object reconstruction without manual segmentation.
Sep 1, 2020
Iteratively weighted principal component analysis and orientation consistency for normal estimation in point cloud
Iteratively re-weighted PCA combined with an orientation consistency propagation step robustly estimates surface normals in noisy and non-uniformly sampled point clouds, outperforming standard PCA normal estimation on sharp features and outlier-contaminated data.
Jan 1, 2020
HLO: Half-kernel Laplacian Operator for surface smoothing
HLO is a half-kernel Laplacian operator for mesh surface smoothing that preserves sharp features and alleviates shrinkage artifacts better than uniform Laplacian denoising, performing on par with or better than state-of-the-art on high-noise meshes.
Jan 1, 2020
Transformation of LEGO Models
Computational method for transforming one LEGO model into another by optimizing brick reuse rate and movement cost, addressing a problem not previously explored in the literature.
Jan 1, 2018
Pick-and-place process sequencing for transformation of rasterized 3D structures
Optimization for robotic pick-and-place process sequencing in 3D rasterized assembly, formulated as a Hungarian-method assignment problem and generalizable from crane operations to large-scale 3D printing.
Jan 1, 2017