Denoising Filter

Additional Material to our submitted CAIP 2017 paper. This is ongoing research work.

Project Contact: Simon Reich

We have benchmarked our algorithm based on the Berkeley Segmentation Dataset and Benchmark and the MS Common Objects in Context Dataset. Due to the huge size of the latter benchmark, we only show Berkeley results here:

Example Images:

Original Noisy Proposed EPF Bilateral Filter Gaussian Filter Median Filter Nonlocalmeans Filter
Gauss Var=10
Gauss Var=30
Gauss Var=50
Gauss Var=100
S&P
 
Gauss Var=10
Gauss Var=30
Gauss Var=50
Gauss Var=100
S&P
 
Gauss Var=10
Gauss Var=30
Gauss Var=50
Gauss Var=100
S&P
 
Gauss Var=10
Gauss Var=30
Gauss Var=50
Gauss Var=100
S&P


© 2011 - 2016 Dept. of Computational Neuroscience • comments to: sreich _at_ gwdg.de • Impressum / Site Info