Noise is often introduced during the image acquisition and image transfer process. The presence of noise affects image detail feature detection. The existing image multi-focus fusion methods could not identify meaningful image features from noise. An antinoise multi-focus image fusion algorithm is presented for image fusion. The improved adaptive block-based image fusion algorithm combined with a new focus measure with noise immunity solves the problem of noisy image fusion effectively and achieves good fusion results. Root Mean Square Error (RMSE) and Mutual Information (MI) are selected to evaluate the fused image with noise of different intensity. Experimental results show that compared with contrast pyramid

wavelet transform and Contourlet transform

the average RMSE of the fused image by the new proposed method is lower 4.0924

4.1268 and 4.0735 respectively

while the average MI is higher 2.0376，2.5875 and 1.3638 respectively. The useful information of the source images can be maintained accurately with the noise interference outperforming other conventional methods in terms of fusion quality and noise reduction in the fused output.