Supplementary Material: Robust 3D Reconstruction via Structured Light Homography and SVD Plane Fitting

This repository contains the full reproducibility package—including source code, simulation frameworks, and raw experimental datasets—associated with the research article: Robust 3D Reconstruction via Structured Light Homography and SVD Plane Fitting. The full stepped vertical distance measurements table is also included, both able to re-obtain and to see it directly. It is also included a direct comparison with Ping and Liu's method from the article «A calibration method for line-structured light system by using sinusoidal fringes and homography matrix». Optik, vol. 261, July 2022, p. 169192. ScienceDirect, https://doi.org/10.1016/j.ijleo.2022.169192. The software implementation provides a robust calibration framework for Line-Structured Light (LSL) systems, utilizing homography mapping and Singular Value Decomposition (SVD) orthogonal consensus to compensate for arbitrary laser plane orientations and mechanical misalignments.The software necessary to obtain the compared method is also provided. Repository Contents: Calibration & Reconstruction Algorithms (MATLAB): Core scripts for the SVD-based "Super Plane" estimation. Homography mapping and feature extraction routines. Iterative pulse-to-millimeter ratio optimization loop. Simulation Framework: Numerical simulation scripts validating geometric integrity under Gaussian noise. Experimental Datasets: Raw calibration data points (Calibration Pattern). Reconstruction point clouds for the stepped artifact, Grade 1 gauge blocks, Aluminum profile, electronic PCBs, silicone hand and automotive components used in the study. Results Extraction: Helper scripts to generate the RMSE error plots and qualitative reconstruction figures presented in the manuscript. Ping and Liu baseline: Python Scripts and data to obtain their Homography. System Requirements: For Matlab MATLAB (Base) Computer Vision Toolbox Image Processing Toolbox Optimization Toolbox Statistics and Machine Learning Toolbox Lidar Toolbox **OR** Robotics System Toolbox (for shared functionalities) For Python Python 3.12.9 numpy==1.19.5 opencv_python==4.13.0.92 Pillow==12.2.0 scipy==1.5.4