The Research Of Image Fusion Algorithm Based On Wavelet Transform

Image fusion is an important part of image processing. It can collaboratemultiple sensor image information of the same scene and output a fused image whichis more suitable for human visual perception or computer to do further process andanalysis. It can significantly improve the lack of a single sensor to improve thesharpness and to increase the amount of information included in the resulting image,and it is helpful to obtain more accurate, reliable and comprehensive information ofthe target or scene. Therefore, the research of image fusion technology is animportant theoretical and applicational issue. With the continuous development andimprovement of multi-sensor image fusion technology, its application in military andcivilian fields will be more extensive. In-depth research on is important in. It isimportant for national economy development and national defense construction toconduct in-depth research of image fusion technology.This paper introduces a basic concepts and theories of image fusion andcompares different fusion algorithms, at last, determines to adopt wavelet transformmethod to do the image fusion. The key of wavelet transform fusion processing is theselection of wavelet parameters. This paper presents the parameter selectionindicators. Then introduce subjective evaluation standards and objectiveevaluation parameters of image quality in detail.In view of shortage of the traditional wavelet contrast fusion algorithm inmulti-focus image fusion, we choose the wavelet neighborhood energy contrastfusion algorithm. The simulation results show that the fusion algorithm is better thanthe traditional wavelet contrast fusion algorithm in evaluation index such asdefinition, average gradient and information entropy.This paper studies the features of the infrared and visible light images andproposes fusion algorithm which combine edge extraction with local variancemethod. The simulation results show that the fusion algorithm fused have a good effect on the subjective evaluation and objective evaluation parameters of images.