Research On Medical Image Fusion Algorithm Based On Multi-scale Morphological Gradient

With the rapid development of science and technology,medical imaging technology is also constantly improving,and the medical image processing methods in clinical diagnosis are becoming more and more diversified.However,in the current medical diagnosis,the single-mode medical image has the defects of relatively single information and poor focus positioning ability.It is difficult to meet the needs of modern medical treatment.Therefore,it can provide more effective information and improve the focus positioning ability.Image fusion technology came into being.It is precisely because of this demand in clinical medicine that image processing researchers gradually focus on the research of medical image fusion.The edge information,detail information and image clarity of medical fusion images play an important role in locating different organs and diseases,and can provide a more adequate basis for disease diagnosis.However,the current medical fusion images will have the problems of edge distortion,blurred contours,loss of details and easy production of artificial textures.Such fusion results cannot provide convenience for clinical diagnosis.In order to improve the problems of medical fusion images,It is proposed that a medical image fusion algorithm based on multi-scale morphological gradients in this case.The non-subsampled shear wave(NSST)is used as a tool for image multi-scale decomposition,and then MSMG is used to process the decomposed high and low frequency information.For low-frequency information processing,an improved guided filtering fusion rule based on MSMG is used.When the image is decomposed by NSST,the low-frequency information is prone to produce artificial texture.MSMG is used to obtain the gradient change of the low-frequency information during the decomposition process.According to this change,the regularization parameter in the guided filtering is corrected,and then the guided filtering is used to smooth the low-frequency information.,Suppress the production of artificial texture.The processing of high-frequency information is to make full use of the ability of MSMG to obtain gradient information,extract the boundary detail information of the image,and retain this information in the fusion image,so that the boundary of the fusion image is clearer and the details are richer.In this case the fusion experiment of CT/MR-T2 grayscale image and MR-T2/SPECT-Tc color image is carried out,and the experiment is compared with several existing fusion algorithms,and the fusion image is evaluated through subjective evaluation and objective evaluation.Experimental results show that the overall visual effect of the fusion image proposed by the algorithm is good,the edge contour is clear,and the detail information is better preserved.At the same time,the EN,AG,EIN and FD indexes of the algorithm in this paper are optimal,and the STD and SF indexes are good.On the whole,the fusion image of the algorithm proposed in this paper has rich information,clear edge and detail information,and can provide a more reliable basis for clinical diagnosis.