Python port of source code for KNN Matting, CVPR 2012 / TPAMI 2013 http://dingzeyu.li/projects/knn/
- python 3.5 (Though it should run on 2.7)
- scipy
- numpy
- matplotlib
- sklearn
- '''python knn_matting.py'''
- mylambda (λ) is a constant controlling the users confidence in the constraints
- image size not larger than 640*480 reccomended for speed and memory reasons.
The sparse matrix solve step may be greatly sped up by using the scikit-umfpack wrapper of UMFPACK sparse direct solver, https://github.com/scikit-umfpack/scikit-umfpack
For more information see the orginal paper and it's website here http://dingzeyu.li/projects/knn/
The original matlab code is here https://github.com/dingzeyuli/knn-matting
The code is free for academic/research purpose. Use at your own risk and we are not responsible for any loss resulting from this code. Feel free to submit pull request for bug fixes.
Original authors:
Qifeng Chen (cqf@stanford.edu) and Dingzeyu Li (dli@cs.columbia.edu)