结合Shearlet变换和果蝇优化算法的甲状腺图像融合

Thyroid Image Fusion Based on Shearlet Transform and Fruit Fly Optimization Algorithm

针对甲状腺肿瘤超声图像复杂度高和SPECT图像边界模糊的特点,结合Shearlet变换能够捕捉图像细节信息和果蝇优化算法可靠性高的优势,提出了Shearlet变换和果蝇优化算法相结合的图像融合算法。首先,用Shearlet变换对已精确配准的源图像进行分解,分别得到高低频子带系数。高频子带系数采用区域能量取大的融合规则,低频子带系数使用改进的加权融合规则,并把果蝇优化算法引入低频融合过程,以互信息作为适应度函数来获取最优值,克服了原加权融合算法互信息低的缺点。最后,用Shearlet逆变换得到融合后的图像。实验结果表明,此算法在主观视觉效果和客观评价指标上优于其他融合算法。

According to the characteristics of ultrasound images with high complexity and SPECT image with blurred boundary, combining the advantage of the Shearlet transform can capture the detail information of images and the high reliability of the Fruit Fly Optimization Algorithm, an image fusion algorithm based on Shearlet trans-form and Fruit Fly Optimization Algorithm is proposed. Firstly, the Shearlet transform is used to decompose the reg-istered source images, thus the low frequency sub-band coefficients and high frequency sub-band coefficients can be obtained. The high frequency sub-band coefficients are fused by the region energy maximum. The fusion rule of the low frequency sub-band coefficients is based on the method of modified weighted fusion, in order to overcome the disadvantage of low mutual information in primary weighted fusion algorithm, the Fruit Fly Optimization Algorithm is introduced in fusion process, the mutual information as fitness function is used to calculate the optimum solution. Finally, the fused image is reconstructed by inverse Shearlet transform. The experimental results demonstrate that the proposed method outperforms the other methods in term of visual evaluation and objective evaluation.