Convolutional Sparse Coding-based Image Decomposition

He Zhang and Vishal Patel

Abstract

We propose a novel sparsity-based method for cartoon and texture decomposition based on Convolutional Sparse Coding (CSC). Our method first learns a set of generic filters that can sparsely represent cartoon and texture type images. Then using these learned filters, we propose a sparsity-based optimization framework to decompose a given image into cartoon and texture components. By working directly on the whole image, the proposed image separation algorithm does not need to divide the image into overlapping patches for leaning local dictionaries. Extensive experiments indicate the proposed method performs favorably compared to state-of-the-art image separation methods.

Session

Posters 2

Files

PDF iconExtended Abstract (PDF, 1M)
PDF iconPaper (PDF, 3M)

DOI

10.5244/C.30.125
https://dx.doi.org/10.5244/C.30.125

Citation

He Zhang and Vishal Patel. Convolutional Sparse Coding-based Image Decomposition. In Richard C. Wilson, Edwin R. Hancock and William A. P. Smith, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 125.1-125.11. BMVA Press, September 2016.

Bibtex

        @inproceedings{BMVC2016_125,
        	title={Convolutional Sparse Coding-based Image Decomposition},
        	author={He Zhang and Vishal Patel},
        	year={2016},
        	month={September},
        	pages={125.1-125.11},
        	articleno={125},
        	numpages={11},
        	booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
        	publisher={BMVA Press},
        	editor={Richard C. Wilson, Edwin R. Hancock and William A. P. Smith},
        	doi={10.5244/C.30.125},
        	isbn={1-901725-59-6},
        	url={https://dx.doi.org/10.5244/C.30.125}
        }