All Projects → sai-bi → L1Flattening

sai-bi / L1Flattening

Licence: MIT license
An L1 Image Transform for Edge-Preserving Smoothing and Scene-Level Intrinsic Decomposition

Programming Languages

C++
36643 projects - #6 most used programming language
matlab
3953 projects
c
50402 projects - #5 most used programming language

#L1 image flattening <img src = "./data/1.png"/ width="45%">

original | flattening

<img src = "./data/2.png"/ width="30%"> <img src = "./data/2-flat.png"/ width="30%"> <img src = "./data/2-smooth.png"/ width="30%"> original | flattening | edge-preserving smoothing

This is the code for our 2015 SIGGRAPH paper An L1 Image Transform for Edge-Preserving Smoothing and Scene-Level Intrinsic Decomposition. If you use this code, please cite our paper.

Please note that this implementation may not necessarily follow exactly what the paper describes. Given the size of energy function to be optimized, it may require 4 GB memory or more. In addition, to generate super-pixels, I have made use of an external program. For details, please see demo.m.

If you need to rebuild the source files under 'src/', please run mex -I.\ segment.cpp.

If you have any question, please contact me at soundsilencebisai-at-gmail-dot-com.

Citation

@article{L1Intrinsic,  
    author = {Sai Bi and Xiaoguang Han and Yizhou Yu}, 
    title = {An $L_1$ Image Transform for Edge-Preserving Smoothing and Scene-Level Intrinsic Decomposition}, 
    journal = {ACM Trans. Graph. (Proc. SIGGRAPH)}, 
    volume = {34}, 
    number = {4}, 
    year = {2015}, 
    publisher = {ACM}, 
}
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].