基于多小波基的医学图像融合

Medical Image Fusion Based on Multiwavelet Base

针对CT图像和MRI图像所显示组织信息的不同,提出一种新的基于多个小波基的图像融合方法。首先利用多个不同的小波基对医学图像CT和MRI进行小波分解,高频采用方向对比度融合规则,低频采用加权平均的融合规则得到相应小波基的融合子图像,最后将各子融合图像采用像素平均的融合方法,得到最终的融合图像。实验结果表明,无论是在视觉效果上还是在定量指标上,该方法融合效果均明显优于单一小波基的图像融合方法。

For the medical CT and NMR imge display different tissue imformation,a multiwavelet base fusion method is proposed. At first,different wavelet base is employed to wavelet transform for CT and NMR,these decomposed high frequency image are fused using direct contrast measure fusion rule,low frequency is fused by average fusion rule. Then, sub - fusion image is fused using pixie average method. At last, image needed are gained, experimental results show that the new method presented here is much effective in image fusion.