Research On The Medical Images Fusion Algorithm Based On Beyond Wavelets Transform

The vigorous development of the medical device,driven by the development of modern science and technology,the medical imaging equipment have also flocked to the emergence of,the medical image more and more.Medical diagnosis,patients need to take different medical images,which requires skilled professional doctors can according to the medical imaging diagnosis of the patient and the doctor at the time of diagnosis need combined with the medical image information,which the doctor especially internship is a great challenge.The medical image fusion technology,which is the principle of the integration of different medical images together,the medical image of the patient's medical image information on a medical image display.This paper mainly on medical image fusion algorithm in the presence of by of wavelet transform,wavelet transform,local entropy based on pulse coupled neural networks(PCNN)and sparse representation and other medical image fusion algorithm for understanding and thinking,two new medical image fusion scheme was put forward.The main innovative work is as follows:1.In order to improve the quality of medical image fusion,we propose a neural network(PCNN)fusion and neighborhood entropy method based on pulse coupled Nonsubsampled Contourlet Transform(NSCT).Taking full account of the characteristics and image NSCT domain coefficients PCNN basis consistent with the visual characteristics of the human eye,for the low,high frequency are proposed based on improved energy and Laplace(Sum-Modified-Laplacian,SML)of low-frequency coefficients Fusion and a fusion method based on neighborhood entropy PCNN high-frequency coefficients,build a new fusion rules used in medical image fusion model.New fusion algorithm can achieve better visual effects and better objective indicator values,verify the superiority of the algorithm.2.The proposed fusion YUV transformation and structural and functional medical images based on sparse representation algorithm.Gradation characteristic information using the color information and reservations YUV conversion function Y component image can represent the source image,combined NSCT transform fusion Y component and structure of the image,while the low-frequency method using sparse expression of the low-frequency sub-band fusion,high frequency are proposed based on local energy.Algorithm taking into account the color features and spatial resolution to enhance two aspects of integration achieved good results.