Skip to content

Latest commit

 

History

History
21 lines (16 loc) · 1.27 KB

README.md

File metadata and controls

21 lines (16 loc) · 1.27 KB
Raw

Image-Inpainting-Sparse-Representation

Image inpainting using sparse representations (Orthogonal Matching Pursuit (OMP) and Iteratively Reweighted Least Squares Method (IRLS)). This was done as a part of the course work for EE698K at IIT Kanpur.

Results

PatchFillingResults
OMPandIRLS results
For more details, please refer to Report.pdf

Steps

  1. Run the SparseInpainting.py using python SparseInpainting.py
  • The parameters (like dictionary size, patch size etc.) can be changed from this file itself. Refer to line #8 to line # 14.
  • Don't forget to change line #73, #75, #76, and #113 if you want to run the algorithm on different files (the default is Daniel Radcliff).
  • Comment/uncomment line #183 and #184 to run OMP or IRLS
  1. Run computeNIQE.py using python computeNIQE.py to get the NIQE value
  • Please change the filename in line #5
  1. Report.pdf is the report for the assignment

Comments/Suggestions

Please feel free to open an issue, in case you have a question/comment/suggestion. I will try to revert as soon as possible.