Open AccessMulti-spectral and panchromatic image fusion based on HIS-wavelet transform and MOPSO algorithm
Author(s) -
Zhao Liaoying,
Ma Qi-Liang,
Xiaorun Li
Publication year - 2012
Publication title -
wuli xuebao
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.199
H-Index - 47
ISSN - 1000-3290
DOI - 10.7498/aps.61.194204
Subject(s) - panchromatic film , multispectral image , image fusion , particle swarm optimization , wavelet , transformation (genetics) , wavelet transform , computer science , artificial intelligence , algorithm , computer vision , pattern recognition (psychology) , image (mathematics) , biochemistry , chemistry , gene
Effective fusion method of remote sensing multispectral and panchromatic image must ensure maximizations of spectrum and space information. Using the fusion algorithm framework with combining HIS transformation and wavelet transform (HIS-wavelet), in this paper we propose a new method to extract high frequency coefficients and a new multispectral and panchromatic image fusion method by using multi-objective particle swarm optimization (MOPSO) algorithm. According to the physical characteristics that the edge information in the high frequency has the nature of direction and noise points in the high frequency are generally isolated, a high frequency coefficient extraction method based on Gauss first order differential is proposed. The final resulting image is optimally combined by two images obtained by using different fusion rules in HIS-wavelet. Multiple fusion evaluation indicators are used as object functions and the MOPSO algorithm is used to find the optimal weights. The experiments on TM multi-spectral image and SPOT panchromatic image are carried out. Experimental results demonstrate that the improved method has a better improvement in spectral and spatial information. At the same time, the resulting image which is obtained using MOPSO algorithm has obvious advantages in retaining the spectral information and the spatial information is also greatly improved.