A Single-shot-per-pose Camera-Projector Calibration System For Imperfect Planar Targets

https://github.com/BingyaoHuang/single-shot-pro-cam-calib

Introduction

A single-shot-per-pose camera-projector calibration method that deals with imperfect planarity of the calibration target.

Highlights:

• The proposed method performs camera-projector pair calibration with only a single-shot-per-pose, making it practically convenient in many applications. It can provide flexible and accurate results even when the board is handheld.
• A bundle adjustment (BA) algorithm is developed to jointly optimize the estimated camera/projector models and feature points in model space.
• Unlike many existing methods, to calibrate the system, we apply points from a Structured Light (SL) pattern rather than from a checkerboard. This strategy boosts not only the number of feature points but also their spatial distribution, and hence improves calibration robustness.

For more info please refer to our paper. To reproduce paper results please refer to ismar18 branch.

Updated 04/03/2019: 3D reconstruction feature released. Click Calibration tab and load a calibration yml file by clicking Load Calibration, then select a set in the Images list and click Reconstruct.

Required Software & Packages

• MATLAB 2018b
• mexOpenCV v3.3.0

Usage

1. Watch a short video tutorial here.
2. Run calibApp.mlapp to start the App.