Conference Proceeding

A Novel Fast Image Fusion Algorithm Based on Directional Contrast and Weighted Activity

Northwestern Polytech. Univ., Xi'an;
06/2007; DOI:10.1109/ICIEA.2007.4318378 ISBN: 978-1-4244-0737-8 pp.102 - 106 In proceeding of: Industrial Electronics and Applications, 2007. ICIEA 2007. 2nd IEEE Conference on
Source: IEEE Xplore

ABSTRACT Image fusion is a technique to obtain integrated information from multiple images of the same scene. Because the complementary information of different images is effectively used and the interference of redundant information is reduced, the fused image contains richer and more accurate description of the scene than any of the individual source images. In this paper, a novel fast image fusion algorithm based on directional contrast and weighted activity is proposed. Firstly, lifting wavelet transform is applied to the source images to obtain the wavelet coefficients. Secondly, different fusion strategies are employed in different frequency bands to construct wavelet coefficients of the fused image. Finally, the fused image is obtained by lifting wavelet reconstruction. The experimental results of multi-focus and multi-spectral images show this newly proposed fusion algorithm can obtain good performance with lower computational cost.

0 0
 · 
0 Bookmarks
 · 
46 Views

Keywords

accurate description
 
complementary information
 
different fusion strategies
 
different images
 
directional contrast
 
experimental results
 
fused image
 
good performance
 
Image fusion
 
individual source images
 
lower computational cost
 
multi-spectral images
 
multiple images
 
novel fast image fusion algorithm
 
proposed fusion algorithm
 
redundant information
 
source images
 
wavelet coefficients
 
wavelet reconstruction
 
weighted activity