Research On Fusion Algorithm Of Multi-exposure Image

Multi-exposure image fusion technology can effectively solve the problem of image feature information loss caused by over-exposure or under-exposure in high dynamic ran ge scene,and is able to more truly record natural scenes.How to preserve clear details of the bright and dark areas in the fused image without distortion,and remove ghost artifacts of exposure fusion for dynamic scenes are the hardest problems of current multi-exposure fusion technology.In view of the above problems,this paper puts forward the multi-exposure fusion algorithm under both static and dynamic scenarios,and the main contents and results are as follows:(1)A multi-exposure image fusion algorithm based on super-pixel segmentation is proposed in this paper.The traditional patch-based fusion methods use experimental tests to set the shape and size of the image patch,which reduces the fusion performance and has poor robustness.Firstly,the source images are divided into perceptual image patches by super-pixel segmentation.Secondly,structural patch decomposition is conducted on the image patches to obtain three individual components,and different fusion rules are designed according to the characteristics of each component.Thirdly,the weight map of each component,signal strength component and brightness component are smoothed by guided filtering,effectively overcoming the problem of block.This method can fully retain the feature information and improve the visual quality of the fused image.The experimental results show that,this method is able to retain more feature information in the input images and no artifacts are produced.(2)In order to solve the ghost problem of dynamic scene,dynamic multi-exposure image fusion based on midway image equalization is proposed in this paper.The exposure level of each input image in multi-exposure image sequence is different,which leads to the difficulty of moving target detection.Firstly,the input images and the reference image are processed by the midway image equalization,and the image pairs with the same exposure degree are obtained.Secondly,the motion target is detected by difference method and the motion weight maps are optimized by morphological processing.The experimental results show that the proposed method can effectively avoid the phenomenon of the ghost,and obtain fused image with rich details and good visual effect.