Research Of Medical Image Fusion Based On Latent Low Rank Decomposition And Dictionary Learning

With the development of human society and the improvement of medical standards,the development of the medical field has received the attention of scholars from all over the world.The constant innovation of new technologies and sciences has enabled all kinds of advanced medical equipment to be used in today's major hospitals.Medical imaging equipment,.usually using different methods to obtain images of the lesions of patients,due to the limitations of equipment and technology,single-modal medical images often can not clearly show the symptoms,need to combine multi-modal medical images,in order to diagnose the disease Consultation.Due to the different techniques used in medical equipment,the acquired medical images show different pathologies,but there is a certain relationship between them,such as the existence of complementary information in the content.Therefore,studying multimodal medical image fusion to obtain more image details is of great significance for clinical medical diagnosis.For medical image fusion directly using sparse representation,it is difficult to obtain better fusion effect.This paper combines medical image fusion technology,low rank representation and sparse representation correlation theory to improve the medical image fusion method of secondary low rank decomposition and sparse representation.A medical image fusion method based on the combination of potential low rank decomposition and sparse dictionary expression is proposed,which hopes to help medical work.The main work and innovations of this paper are as follows1.Because the method based on quadratic low rank decomposition and sparse representation fusion,the information is too much missing and the calculation is complicated.An improved method is proposed:a low rank decomposition is performed in the decomposition stage,the gray value"absolute value"is used to enlarge the rule for the low rank component;the overcomplete dictionary sample selection is not related to the medical field.Image.The improved method reduces the lack of excessive information,the calculation is small,the time is short,and the quality of the fused image is improved.2.Based on low rank decomposition and sparse combination,a medical image fusion method based on latent low rank decomposition and dictionary learning is proposed.The source image to be fused is decomposed into a low rank component,a sparse component,and a sparse noise component by a potential low rank decomposition method.The gray value"absolute value"is used to enlarge the rule for the low rank component;for the sparse component,the potential low rank decomposition model is designed,the image set is decomposed,and the K-SVD algorithm is used to train the low rank dictionary and the sparse dictionary respectively.For sparse reconstruction;eliminate sparse noise.In order to verify that the proposed method can be applied to the medical field,the experimental simulation and analysis are carried out.The results show that compared with the other seven sparse fusion methods,the proposed method has significant visual effects and good image fusion quality.