Multisource Images Adaptive Fusion Based on Double Density Dual-tree Complex Wavelet Transform

Abstract

Double density dual-tree complex wavelet transform (DD-DTCWT) as multiscale decomposition tool provides multiresolution and multidirection expansion for images. Compared with the discrete wavelet transform (DWT), it is shift-invariant and can overcome the detail blur and pseudo-Gibbs phenomena. Fusion rule is the key that influences the quality of image fusion. This paper propose a fusion rule based on improved pixel weighted average(IPWA) in approximate images and coefficient absolute value with neighborhood average energy consistency selection(ANECS) in detail images. The proposed fusion rule is associated with DD-DTCWT to make full use of the characteristics of them. Composite coefficients is reconstructed by double-density dual-tree complex wavelet inverse transform (DD-DTCWIT). The performance of the proposed method in this paper is compared to a few existing techniques in the literature. The experiment results demonstrate that the proposed method performs better than those in the literature. In addition, it also reveals that the proposed technique is more robust than those in the literature.