The Performance Analysis of Image Fusion Algorithm

Author

Zhao, Chunling ; Ding, Quanxin ; Li, Jianxun

Author_Institution

Key Lab. of Fire Control Technol., Luoyang

Volume

2

fYear

2008

fDate

17-18 Oct. 2008

Firstpage

83

Lastpage

86

Abstract

The property of Multiscale transform can be used to decompose a complex and highly correlated problem into a set of irrelevant or less correlated problems, which is very useful for processing the information of images for the purpose of fusion. In this paper, two representative multiscale image fusion methods, namely the Laplacian pyramid(LP) and the discrete wavelet transform(DWT), have been studied. The study is focused on the influence of the number of level, fusion rules and wavelet bases to the fusion performance for multi-focus images fusion, since this type of studies was seldom proposed in previous research. The DWT is generally thought more powerful than the LP but the comparison of the fusion performance indicates that LP outperforms DWT.

Keywords

discrete wavelet transforms; image fusion; Laplacian pyramid; discrete wavelet transform; image fusion algorithm; multiscale transform; Algorithm design and analysis; Biomedical optical imaging; Computational intelligence; Discrete wavelet transforms; Focusing; Gaussian processes; Image fusion; Image storage; Layout; Performance analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Design, 2008. ISCID '08. International Symposium on

Conference_Location

Wuhan

Print_ISBN

978-0-7695-3311-7

Type

conf

DOI

10.1109/ISCID.2008.126

Filename

4725462