结合Curvelet变换和LSWT的多聚焦图像融合算法

A multi-focus image fusion algorithm based on Curvelet transform and LSWT

针对多聚焦图像,提出了一种结合二代Curvelet变换和提升静态小波变换LSWT的图像融合算法。首先将待融合的图像分别进行离散Curvelet分解变换,得到不同分解级数和方向下的细节尺度系数和粗尺度系数;其次对粗尺度系数分别进行LSWT变换,对变换得到的低频分量和高频分量分别采用不同的方法融合后进行LSWT逆变换,得到的系数作为Curvelet变换的粗尺度系数;对于Curvelet变换后得到的细节尺度系数采用局部平均能量方差的方法进行融合;最后进行Curvelet逆变换得到融合后的图像。实验结果显示,该方法融合效果较好,优于传统方法。

Focusing on multi-focus images, in this paper we present an image fusion method based on the second generation Curvelet transform and the lifting stationary wavelet transform （LSWT）. Firstly, the images to be fused are decomposed by discrete Curvelet transform, thus the fine-scale and coarse- scale coefficients are obtained in different scales and directions. Secondly, the coarse scale coefficients are decomposed by the LSWT. The low-frequency coefficients and the high-frequency coefficients are sepa- rately fused by different methods. Subsequently the coefficients obtained by the lifting stationary wavelet inverse transform are the coarse-scale coefficients of the Curvelet inverse transform. The fine-scale coefficients are fused by the local average energy and variance method. Finally,the fused image is obtained by the Curvelet inverse transform. Experimental results show that the proposed method is superior to the traditional methods.