Presentation 2014/9/11
A Study of Image Coding Using Multiple Bases Sets Designed under Different Sparsity Condition
Takahiro Imaizumi, Ji WANG, Yoshiyuki YASHIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We had proposed an image coding method using the multiple sparse bases sets based on local features calculated by DSIFT, in which an image to be encoded is divided into blocks and class number of each block is determined, and the sparse bases set designed for the class is utilized for encoding. In this paper, we introduce a new technique using the multiple sparse bases sets designed under different sparsity conditions. In the proposed method, for each class classified with local features, several sparse bases sets are designed under different K-SVD constraint condition about the number of non-zero coefficients, and they are adaptively used block by block in the encoder. Experimental results show that the proposed method can improve R-D coding efficiency compared to conventional one using single bases set for each class.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Image coding / Sparse coding / K-SVD / OMP / DSIFT
Paper # LOIS2014-25
Date of Issue

Conference Information
Committee LOIS
Conference Date 2014/9/11(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Life Intelligence and Office Information Systems (LOIS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Study of Image Coding Using Multiple Bases Sets Designed under Different Sparsity Condition
Sub Title (in English)
Keyword(1) Image coding
Keyword(2) Sparse coding
Keyword(3) K-SVD
Keyword(4) OMP
Keyword(5) DSIFT
1st Author's Name Takahiro Imaizumi
1st Author's Affiliation Graduate School of Information Science, Chiba Institute of Technology()
2nd Author's Name Ji WANG
2nd Author's Affiliation Graduate School of Information Science, Chiba Institute of Technology
3rd Author's Name Yoshiyuki YASHIMA
3rd Author's Affiliation Graduate School of Information Science, Chiba Institute of Technology
Date 2014/9/11
Paper # LOIS2014-25
Volume (vol) vol.114
Number (no) 220
Page pp.pp.-
#Pages 6
Date of Issue