Multi-focus Image Fusion Algorithm Based On Shearlet Transform

Due to the limited depth of field of the optical lens,it is often difficult to clearly focus on an image in different scenes within the same scene.This problem can be solved by image fusion technology.This technology can fuse the important information in multiple images into one image,so as to obtain a more comprehensive and accurate information description of the scene for further image analysis and processing.In recent years,multi-focus image fusion has been widely used in many fields such as information security,medical treatment,military operations,and weather forecasting.Aiming at the problems of low contrast,missing information of edge detail,and failure to adapt to human visual characteristics in fusion images obtained by traditional multi-focus image fusion methods,multi-scale geometric analysis tools are used at the image pixel level to fuse multi-focus images.In this paper,combining the own characteristics and regional characteristics of multi-focus images,the following two algorithms are proposed:1.A new multifocus image fusion method based on Shearlet transform and adaptive Pulse Coupled Neural Network?PCNN?is proposed.Firstly the two source images are decomposed respectively by Shearlet transform to obtain a low-frequency subband and a series of high-frequency subbands with different directions in different scales respectively.Since the sum of the low frequency subbands of the source image contains energy information,the absolute value of difference of the low frequency subbands contains edge texture information.In the proposed algorithm,the two parts are weighted and summed to obtain the fused low-frequency subband,and the weights are calculated from the average gradient of low-frequency subband.The high-frequency subbands of the source image are fused with the adaptive PCNN fusion rules,and the sum-modified Laplacian?SML?are chosen as the excitation for PCNN,and the link strength for PCNN is adaptively calculated from the spatial frequency?SF?of the source image.The fused high-frequency subbands are obtained from the ignition map of the PCNN.Finally,the Shearlet inverse transform is used to obtain the fused image.A large number of experimental results show that compared with the traditional methods,the proposed method has a good effect both in subjective visual and objective evaluation.2.A new multi-focus image fusion method based on Shearlet transform and Particle Swarm Optimization?PSO?algorithm is proposed.Firstly the two source images are decomposed respectively by Shearlet transform to obtain a low-frequency subband and a series of high-frequency subbands with different directions in different scales respectively.For the low-frequency subband,different weighting rules are chosen by comparing the corresponding low frequency coefficients to obtain the fused low-frequency subband,and the optimal weights are found by the PSO algorithm.The fitness function of the PSO algorithm is designed as the product of the Mutual Information?MI?and Q^{AB F}of the low-frequency fused image;For the high-frequency subbands,the weighted New sum-modified Laplacian?NSML?of the region is compared to chosen the corresponding high frequency coefficients to obtain the fused high frequency subbands.Finally,a fused image is obtained by Shearlet inverse transform.A large number of experimental results show that compared with the traditional methods,the proposed method has a good effect both in subjective visual and objective evaluation.