This repo is used as a record of developing a QFR3DReconstructionAPP at Escope Tech from May 2019 to Aug 2019
Contributions:
- Modified the binocular stereo vision method to detect the severity of vascular stenosis using DSA images in Python and C++
- Reconstructed central lines for blood vessels using image preprocessing and epipolar constraint principal
- Acquired 3D cloud point of vessel surface using Pinhole model and Newton-Raphson method
Techniques:
- Algorithms/Models: Epipolar Constraint, Pinhole Model, Newton-Raphson Method
- Frameworks/Libraries: OpenCV, Qt, Spyder, ParaView
- Languages: C++, python
The code and note(in Chinese) are finished on 7/29/2019 which is not the final version of this private project. But you are welcome to contact me and ask any questions about this project.