PET-SPECT And MRI Image Fusion Algorithm Based On Gradient Domain

The development of medical imaging technology provides a visual image for clinical diagnosis and it has been widely used.However,single-modal medical images generally do not provide more comprehensive medical information.How to efficiently integrate multimodal medical imaging information and obtain more comprehensive,accurate and rapid clinical diagnostic images has become an urgent problem to be solved.In order to improve the sharpness and visual effect of images efficiently,this paper proposes a medical image enhancement technique based on the gradient domain.And,combining it with the multi-scale decomposition scheme of Laplacian pyramid,the paper proposes a multi-scale medical image fusion algorithm based on gradient enhancement.The main research work is as follows:1)First,the background knowledge and research significance of image processing are briefly introduced.Then,the article explains the fusion algorithm and evaluation method from the perspective of image fusion,emphasizing the characteristics and research significance of medical images.Finally,the paper introduces the image enhancement algorithm and multi-scale fusion algorithm,and summarizes the advantages and disadvantages of the algorithm and the improvement direction.2)Aiming at the Medical images demand high on contrast and texture,this paper proposes a new image enhancement algorithm based on gradient domain.Firstly,the image contrast and texture feature extraction are performed on the input medical image,and two reference functions for the image detail region and the edge region are respectively designed by using the feature information.Then,after fully analyzing the correlation between reference functions,this paper proposes an adaptive weighting coefficient to be used as an important reference for gradient mapping.Finally,the gradient coefficients are multiplied by the original image to obtain an enhanced image.Because the gradient lifting algorithm is sensitive to image texture and is not sensitive to image noise,the algorithm can effectively suppress image noise while improving image contrast and texture,so a clearer enhanced image can be obtained.3)This paper proposes a multi-scale medical image fusion algorithm based on gradient domain.Firstly,this paper adopts the Laplacian pyramid decomposition scheme to obtain multi-scale residual sub-band image(RSI)and base sub-band image(BSI).Then,the decomposed RSI image and the BSI image are respectively adopted:an image fusion rule based on interest information and a fusion rule based on a local energy maximum scheme.Finally,the fused image is reconstructed by the Laplacian pyramid inverse operation.The fusion results improve image texture while enhancing image contrast,which makes the fused image clearer.