采用shearlet变换的多聚焦图像融合

相对传统多尺度分析工具，shearlet变换更适于提取图像细节信息。采用shearlet变换进行图像融合，对源图像进行shearlet域分解，对低频子带采用SML算子作为融合依据，高频子带采取区域能量与单个像素相结合的方式选择系数，对融合后的系数进行逆shearlet变换得到融合图像。仿真实验表明，算法在视觉效果和量化结果上均有提高。

Multi-focus image fusion using shearlet transform

Compared with traditional multi-scale analysis method, shearlet transform is more suitable for extracting details of the image. In the paper shearlet is used in multi-focus image fusion. The source images are decomposed into several subbands using shearlet. According to the characteristics of multi-focus images, the coefficients of low-frequency subband are fused with a scheme based on the SML operator. The coefficients of high-frequency subbands are fused with the fusion rule based on both local energy and single pixel. The fused image is obtained by performing the inverse shearlet transform on the combined coefficients. The experimental results show that this method obtains better fusion quality in terms of both visual and quantified measure.