Image fusion has been developed as a powerful method in medical purposes like diagnosis of vascular disease, tumor detection, surgical navigation and planning of treatment for the diseases. In this paper, NSST decomposition based new multimodal medical image fusion (MMIF) method using weighted local energy motivated parameter-adaptive PCNN is proposed. The input images are decomposed into corresponding high and low frequency bands by using NSST. The low frequency coefficients are merged using max rule. The high frequency coefficients are fused by weighted energy motivated PAPCNN. Using inverse NSST, fused image is obtained. The quality metrics, such as standard deviation (STD), information theory based metrics QTE, QNCIE, QMI, and image feature based metrics QP, QG are used to verify the quality of the fused image and to evaluate the performance of proposed method.
Volume 12 | 03-Special Issue
Pages: 960-967
DOI: 10.5373/JARDCS/V12SP3/20201340