Great job!

[HHIR]

it's so interesting :)
i hope you will finish soon as possible

[kuan-wang]

Thanks :)
This repo is just the draft for this project, and the newest version is in my another repo opencv_contrib(branch gsoc_opencv). The final version will be part of OpenCV soon, if you want to try this moudle now, you can git clone https://github.com/THUKey/opencv_contrib.git and build it with opencv.

[kuan-wang]

remember check out to branch gsoc_opencv :)

[HHIR]

I had finished build everything well.
but I don't have Depthsuperres, Latticefilter ... after building.
I just have the bilater file that I don't know how can I test it!
I will be happy if you give me more information. thanks

[kuan-wang]

you can refer to the BilateralSolver.cpp of the repo The_Bilateral_Solver to build the depth demo with bilateralsolver. And the Lattice filter is not a part of this OpenCV moudle, you can refer to the permutohedral_lattice.cpp or the Adams' paper webpages(http://graphics.stanford.edu/papers/permutohedral/) to build it. By the way, you can try the colorization sample(opencv/build/bin/example_ximgproc_colorize)and the disparity filtering sample(opencv/build/bin/example_ximgproc_disparity_filtering) if you have built opencv_contrib with opencv :)

[anilrgukt]

Hi all,

Below mentioned incorporation could make this more useful for the general audience.

1. Can you provide a bit documentation on opencv python interface for this?

2. Also, it would be great if you could implement the forward pass and back-propagation in python required for using this solver with deep neural networks, as is also suggested by the authors in the paper.

Thanks,

[flow-specter]

Hi, may I ask that how you get the confidence image?

[1ros0]

For just stereo (not depth super resolution) are the steps any different?