摘要
由于离散小波变换在图像融合中的缺陷,提出基于双树复小波变换的图像融合方法。首先对源图像进行双树复小波变换,对低频和高频分别采用局部窗口能量比和局部梯度比作为融合相关性测度,然后基于加权和选择相结合的规则融合高频系数和低频系数,得到各频域系数,最后复原图像。实验结果表明,双树复小波变换的融合方法具有良好的客观评价性能和主观视觉效果,融合规则的可行性。
As the defects of discrete wavelet transform in image fusion, We proposed a image fusion method which is based on the Dual-Tree Complex Wavelet transform. First, we should take Dual-Tree Complex Wavelet Transform to the source image, with regard to low-frequency and high-frequency, we take Local window energy ratio and the local gradient ratio respectively as a measure of convergence-related. Then, based on the rules that weighted combined with the options, we integrate high and low frequency coefficients to get the frequency-domain coefficients. At last, we restore the image. The results of experiment show that the fusion method of dual-tree complex wavelet transform has a good performance of subjective evaluation, good visual effects of objective and The feasibility of fusion rule .
出处
《电子测试》
2010年第6期32-36,共5页
Electronic Test
