Source Code for MeshFlow Video Denoising
OpenCV Version:3.1.0
This source code may not have the same processing speed as the paper illustrated. Because I replace the original matching algorithm into a simple one. This may also influence the quality of the denoising result a little. For more project detail, including the academic paper, project abstract and test data, please visiting the project page at http://www.liushuaicheng.org/ICIP/2017/index.html.
- Create a new folder
build - Inside the folder, using the code below for cmake to build the project files:
cmake -G "Visual Studio 12 Win64" DCMAKE_BUILD_TYPE=Release ..If your IDE is Visual Studio 2015, Using the line below to compile your project:
cmake -G "Visual Studio 14 Win64" DCMAKE_BUILD_TYPE=Release ..- Move the test video into this new folder and run this project.
For Visual Studio users, when run the project file, you should change the single startup project option in the solution property into the correct one (not the ALLBUILD.EXE).
For Windows users, using Cmakelists.txt as the cmakelist. (Remember to change the path for the OpenCV build folder.)
For Linux users, using Cmakelists_Linux.txt as the cmakelist.
For further questions, feel free to contact me at albuspeter.rzh@gmail.com.
Thanks Guo Heng who helped me on the Cmake lists.
If you find this useful in your research, please cite our paper "Meshflow Video Denoising" (PDF):
@inproceedings{ren2017meshflow,
title={Meshflow video denoising},
author={Ren, Zhihang and Li, Jiajia and Liu, Shuaicheng and Zeng, Bing},
booktitle={2017 IEEE International Conference on Image Processing (ICIP)},
pages={2966--2970},
year={2017},
organization={IEEE}
}