滤波器组合对医学图像融合效果的影响

滤波器的不同组合影响医学图像融合性能,恰当地进行滤波器组合是医学图像融合技术的前提。本文通过优化滤波器组合对无下采样Contourlet变换(nonsubsampled contourlet transform,NSCT)医学图像融合影响进行深入讨论,比较了不同滤波器组合后的融合结果。结果显示滤波器组合对融合性能有较大影响,lax滤波器组合图像融合效果各项指标分别优于db3滤波器组合。实验结果证明,使用合适的滤波器组合,即使在融合准则很简单、分解层数较少的情况下,同样可以获得很好的融合结果,从而大大减少融合算法的复杂度。

Effect of Filters Combined on Medical Image Fusion

The filters combined influence the qualities of medical image fusion. It is important premise to select filters correctly for Medical Image Fusion Technology. Effects of the nonsubsampled contourlet transform（NSCT） on image fusion were analyzed and fusion results using different filters were discussed in depth. The results confirmed that the filters combined significantly impacted on the qualities of fusion, and the indicators of lax combined filter were better than db3 combined filter. Experimental results showed that choosing appropriate filters combined could achieve a satisfying fusion, even with an easier fusion rule and less decomposing levels, which could reduce the fusion algorithms complexity.