3D reconstruction algorithms, geometry systems, and production engineering.

About Me

I am a 3D reconstruction algorithm engineer working across photogrammetry, multi-view geometry, point clouds, meshes, and production reconstruction systems.

My work combines algorithm design, C++/CUDA implementation, validation, and performance optimization, with an emphasis on large real-world scenes and reliable engineering delivery.

我专注于三维重建、摄影测量与多视几何方向,长期从事 SfM/MVS、SLAM、点云、网格、纹理与大规模重建系统研发。

工作覆盖算法设计、工程实现、性能优化和质量验证,关注真实业务场景中的精度、效率、稳定性与可交付性。

Current focus. 3D Gaussian Splatting, end-to-end 3D reconstruction, and foundation-model-assisted geometry pipelines.

Focus Areas

Reconstruction

SfM, MVS, SLAM, bundle adjustment, panoramic geometry, aerial and indoor reconstruction.

Geometry Processing

Point-cloud filtering, registration, mesh reconstruction, structural modeling, and texture mapping.

Production Systems

Large-scene tiling, parallel pipelines, memory-bounded processing, LOD, and browser delivery.

Engineering Practice

C++/CUDA implementation, numerical validation, performance profiling, and reconstruction quality evaluation.

Selected Projects

Project scope, contribution, and technical stack

Large-scale reconstruction pipeline screenshot

Indoor Large-Scale Reconstruction

Indoor large scenes · Station-based scanning · Cloud reconstruction

Built the indoor large-scene reconstruction system behind station-scanned spaces such as Beike VR house tours, museums, and large venue digitization.

  • Designed the full tiled parallel algorithm framework so reconstruction, mesh processing, texturing, simplification, and LOD generation can run beyond single-machine memory limits.
  • Implemented topology-consistent block mesh merging, robust hole filling, irregular multi-floor segmentation, automatic ceiling hiding, semi-global color balancing, textured simplification, LOD strategy, and cloud Docker deployment.

Tech: C++, point clouds, mesh reconstruction, topology repair, texture optimization, LOD, Docker

Point-cloud colorization visibility and projection diagram

Large-Scene Point Cloud Colorization

Sparse point clouds · Multi-view projection · Semi-global color consistency

Developed a direct colorization pipeline for large sparse point clouds, mapping calibrated multi-view images onto LiDAR geometry while controlling visibility, view selection, and color seams.

  • Built a block-based voxel visibility framework that uses neighboring blocks as occlusion context, Top-K candidate views, point-level projection checks, normal filtering, masks, and rolling-shutter row poses.
  • Designed a distance-driven main-view selection strategy that forms spatially coherent view regions, then applies boundary multi-view consensus and sparse correction-field optimization for semi-global color harmonization.

Tech: C++, OpenVDB, point clouds, ray visibility, multi-view projection, Delaunay graph, conjugate gradient

Handheld LiDAR point-cloud processing pipeline diagram

Handheld LiDAR Point-Cloud Processing

Handheld LiDAR · Point-cloud filtering · Block-parallel refinement

Built a production point-cloud filtering and geometric refinement pipeline for handheld laser scanning, turning raw keyframe clouds into clean global point clouds for downstream reconstruction.

  • Designed a sparse-voxel, block-parallel processing framework with streaming partitioning, neighbor topology, boundary-safe ownership, and global result merging.
  • Implemented time-aware dynamic-object removal, sub-voxel centroiding, robust normal orientation, anisotropic bilateral filtering, MST normal propagation, planar density homogenization, and sparse outlier cleanup.

Tech: C++, OpenVDB, handheld LiDAR, ray traversal, normal estimation, bilateral filtering, Poisson-disk sampling

Indoor structured reconstruction result screenshot

Indoor Structured Reconstruction

Indoor mesh · Structural information · Regularized mesh

Designed a geometry pipeline that recovers room structure from a given indoor mesh and reconstructs a regularized structural mesh.

  • Built the full framework from scratch: space partitioning into a cell complex, visibility-based inside/outside reasoning, and graph-cut optimization.
  • Independently implemented the complex cell-complex geometry algorithms, room partitioning, door recognition, and structure reconstruction modules.

Tech: C++, mesh processing, cell complex, visibility, graph cut, structural reconstruction

Cyclops model refinement before and after screenshot

Cyclops Model Refinement

Panoramic depth · Production geometry cleanup

Refined monocular panoramic depth outputs into more stable point-cloud and mesh geometry for downstream reconstruction pipelines.

  • Added pose consistency checks, outlier filtering, point-cloud fusion, and mesh cleanup.
  • Bridged learned depth predictions with production geometry quality requirements.

Tech: panoramic depth, pose BA, point-cloud fusion, mesh cleanup

CAD-to-floorplan region graph screenshot

CAD-to-Floorplan Parsing

DXF/DWG parsing · Floorplan standardization · Legacy CAD data

Designed an automatic DXF/DWG parser to convert non-standard manually drawn CAD floorplans into a unified structured floorplan representation.

  • Handled inconsistent layers, drawing styles, line types, units, blocks, and geometry conventions across legacy CAD files.
  • Independently designed and implemented algorithms for layer parsing, inside/outside classification, room partitioning, door/window recognition, door block recognition, furniture filtering, and floorplan data generation.

Tech: DXF/DWG, computational geometry, triangulation, region graph, door/window recognition, floorplan JSON

Panoramic geometry preprocessing screenshot

Panoramic Geometry Preprocessing

Single panorama · Spherical image geometry

Developed panorama preprocessing modules for gravity alignment and Manhattan direction estimation before reconstruction and spatial understanding.

  • Implemented cube projection, spherical line reasoning, RANSAC, vanishing-point voting, and fast Manhattan scoring.
  • Packaged geometric priors as reusable preprocessing steps for panorama-based workflows.

Tech: equirectangular panoramas, spherical LUT, vanishing points, gravity rectification, RANSAC

Software

Open tools and engineering outputs

Experience & Education

3D Reconstruction Algorithm Architect
RealSee (贝壳-如视) · Beijing

3D Reconstruction Algorithm Engineer
Feima Robotics (飞马机器人) · Beijing

Co-founder · 技术总监
Infinite3D (北京无限界/景致三维) · Beijing

3D Reconstruction Team Lead
Xi'an Meihang Technology Development Research Institute (西安煤航技术发展研究院) · Xi'an

WebGIS Engineer
Heilongjiang Bureau of Surveying, Mapping & Geoinformation (黑龙江省测绘与地理信息局) · Harbin

M.S. Geographic Information Systems
Wuhan University (武汉大学) · Hubei

B.S. Resource Environment & Urban-Rural Planning
Northeast Petroleum University (东北石油大学) · Heilongjiang

Contact

For technical discussions, open-source collaboration, or reconstruction engineering work, please contact me by email.

huyang0909@gmail.com · github.com/huluoboge