Research And Application Of Infrared Image And Visible Image Fusion

Two types of image fusion refer to the fusion of image content information obtained by two types of sensors to the same scene to obtain an image with two image features.The fusion of infrared images and visible images has many important applications in the military.As an important technology of image fusion,this kind of fusion technology has attracted the attention and rapid development of many researchers.Although the principle and manner of acquiring images by infrared images and visible images are different,the information of infrared images and visible images can be complementary.Infrared image is the sensor obtains image information through the temperature of the target scene.The temperature and emission frequency of the target will affect imaging effect.Infrared image imaging process is not affected by lighting conditions,and can work 24 hours a day.From the visual effect,the thermal target in the infrared image is very significant,but the resolution is low and the sharpness is poor.On the contrary,the resolution of the visible image is higher,and the description of the scene is more specific.However,the imaging process is easily affected by the lighting conditions and does not have the ability to work 24 hours a day.The purpose of this paper is to find a fusion method of infrared image and visible image.By fusion of the source image,a fusion image with obvious target,clear background and good visual effect.The results facilitate the subsequent application in many aspects such as military field and resource monitoring.Based on the research of the basic theory of image fusion,this paper analyzes two kinds of images and proposes a fusion method of infrared image and visible image based on RPCA and NSCT.The main content of this paper can be divided into the three parts:(1)Mainly introduces the image fusion concept and the preprocessing process of image fusion.This paper first introduces the research background and current situation of image fusion,summarizes the existing algorithms and raises the existing problems.Then,after a brief introduction to the image fusion concept and the fusion process,the image denoising,image enhancement,image registration,etc.In the image preprocessing process are introduced in detail,and the working principle and representative algorithm of the preprocessing are introduced in detail.Finally,this paper introduces the existing evaluation methods of image fusion results from subj ective and obj ective perspectives.(2)Mainly introduces the basic theory of fusion and traditional image fusion algorithm.This paper introduces the typical algorithm principle and fusion process in the common algorithm of image fusion.The algorithm is tested by different types of data,and the fusion results of all algorithms are evaluated from two aspects:subjective visual and objective quantitative indicators.(3)Introduce the image fusion algorithm proposed in this paper.Based on the RPCA model,this paper proposes the decomposition of infrared image and visible image,and the obtained low rank component is treated as background information,and the sparse component is treated as the significant target of the image.After comparing the processing effects of several common multi-scale models,this paper uses the fusion algorithm based on NSCT transform to fuse the low-rank components of the image,uses the fusion rule based on region energy to fuse the sparse components,and finally uses the sparse components and low-rank components obtained by fusion.Restore the fusion results.In this paper,three different types of data are merged.It is verified that the fusion effect of the algorithm has advantages in subjective visual and objective evaluation.