Research On Image Fusion Method Based On Multi-scale Decomposition

Image plays a significant role in the operation of various fields in contemporary society.Today,the application of image has become more prevalent due to the development of each imaging technology.Although different imaging technologies have unique characteristics and advantages,there are also some problems that make the obtained image is difficult to meet people's requirements.For example,due to the depth of field of the camera,the imaging is blurred when the object is in the unfocused region,and the commonly used single-mode medical image has a single feature and less information.To solve the above problems,image fusion technology came into being.Multi-focus image fusion and multi-modality medical image fusion are branches of image fusion technology.Multi-focus fusion technology uses algorithms to fuse multiple images with different degrees of focus in the same scene,and finally obtains a fused image,in which all regions and objects are in focus.Multi-modality medical image fusion technology is to integrate medical imaging information in different modalities to finally obtain a comprehensive image of coexistence of morphology and function,which can assist doctors in making a diagnosis of a patient's condition.This paper proposed a fast image decomposition method based on multi-scale theory,which is applied to multi-focus natural image fusion and multi-modal medical image fusion.The specific research contents are as follows:1.A multi-scale image decomposition method based on multi-scale Hessian Matrices(MSH)based on integral graph is proposed for fast decomposition of source images to ensure the timeliness of the fusion method.The multi-scale decomposition method is used to decompose multi-focus images.Firstly,a decomposition strategy is proposed to decompose the source images into feature regions and background regions.Secondly,by designing focus measure in different regions and implementing different fusion strategies,a fused image is obtained.The method proposed in this paper can effectively avoid false imaging in the fused image and solve the problem of false edges.In terms of visual perception and objective evaluation,the proposed method is superior to the latest multi-focus image fusion method.2.Firstly,the fast image decomposition method based on MSH is used to obtain medical image decomposition results at multiple scales,and the multi-scale feature regions and background regions are obtained.Secondly,multi-scale weight maps of different regions are obtained by using a guided filters.Then,by comparing the correlation of feature information between the source images and the decomposition results at different scales,the decomposition results at the optimal scales are selected and used as the weight maps to fuse the source images.Finally,a fused image with multi-modal information is obtained.Experimental results show that the fusion method achieves better results in subjective visual effects and objective evaluation than the existing methods.