Abstract:
An effective multi-focus color image fusion scheme is proposed based on nonsampled shearlet transform (NSST), the sum-modified Laplacian (SML), and combined with dual channel spiking cortical model (DCSCM). Firstly, the source images were translated into HSI color space to obtain the hue (H), saturation (S) and intensity (I). Then, NSST is utilized to decompose the intensity map and obtain its low and high frequency subbands. Secondly, the low frequency subbands are fused by the SML, the high frequency subbands are fused by the DCSCM. Next, NSST inverse transform is performed to obtain the fused intensity map. Finally, according to the Euclidean distance relationship between the source image and the fused intensity map to assign values for the color and saturation, the final fused image is obtained by the inverse HSI. The experimental results show that the proposed method achieves the best visual effect and has advantages in objective evaluation metrics such as mutual information, structural similarity and edge retention.