Generator-of-stereo-speckle-images-with-displacement-labels

A stereo speckle images generator with nonlinear full-field displacement based on Newton iteration. Used for the research of 3D-DIC or other scene flow estimation tasks.

Introduction

• A stereo speckle images generator based on stereo vision theory
• With nonlinear 3D displacement labels, time-wise optical flow labels and right-to-left disparity labels.
• Digit twin of an arbitrary undistorted stereo imaging system(with calibration files)
• Also can be used to generate speckle image paires with pixel displacement, i.e., the left reference and the left current images with the optical flow label.
• For algorithm details and the principles, please refer to ***(unpublished work, update later)

Requests

• python38
• opencv-python (4.4.0 used)
• numpy (1.22.1 used)
• pytorch and torchvision with cuda (torch1.9.0+torchvision0.10.0+cu111)
• (optional) numba

Workflow

• (Optional) Give your stereo calibration parameters and other settings in ./Seeds/input.py
• (Optional) Generate your own seeds by running input.py
• Run main_box.py

Sample

• Dataset sample can be download from Google Drive

• The speckle images sample:

• The 3D displacement labels and the results calculated using 3D-DIC (mm)

• The optical flow and disparity labels and that calculated using 2D-DIC (Pixels):

Cite this work

@article{WANG2022107184,
title = {StrainNet-3D: Real-time and robust 3-dimensional speckle image correlation using deep learning},
journal = {Optics and Lasers in Engineering},
volume = {158},
pages = {107184},
year = {2022},
issn = {0143-8166},
doi = {https://doi.org/10.1016/j.optlaseng.2022.107184},
url = {https://www.sciencedirect.com/science/article/pii/S0143816622002378},
author = {Guowen Wang and Laibin Zhang and Xuefeng Yao}
}