摘要
提出了一种基于非下采样Contourlet变换(NSCT)的红外与可见光图像的融合算法。采用对低频系数取平均,对高频系数中最大分解尺度选择系数最大值,其他尺度系数采用局部方差最大的规则,通过对所得到的融合系数进行逆变换即可得到融合后的图像。实验表明:该算法结合了NSCT的多尺度、多方向和平移不变性的优点,能够更好地提取源图像特征,增强融合图像的空间细节表现能力。融合后的图像具有较好的主观视觉效果,标准差和熵值较传统的融合方法有所提高。
An infrared and visible image fusion algorithm based on the non-subsampled Contourlet transform （NSCT） is provided. Getting the mean of the low frequency coefficients, choosing the maximum value of the highest level＇s coefficients from the high frequency coefficients, applying the local variance maximum principle to other resolution level＇ s coefficients, the fused image after the inverse transforming the fusion coefficients can be thereby got. The experiment indicates that the fusion algorithm can extract the original image features better. The fused image＇ s representation capacity in spatial detail information is also improved, via combing the advantages of the multi-resolution, multi-direction and translation invariancc of the NSCT. Compared with the traditional fusion algorithms, the fusion algorithm provides better subjective visual effect, and the standard deviation and entropy value would be somewhat increased.
出处
《传感器与微系统》
CSCD
北大核心
2008年第12期45-47,共3页
Transducer and Microsystem Technologies
