Research On B Ultrasonic Image Denoising Based On Variation And Partial Differential Equations

Variational partial differential equation has become an important method in the field of computer vision and image analysis.Due to the limitation of the ultrasonic imaging equipment and various external factors,the medical image of B ultrasonic is easily disturbed,and the problem of speckle noise appears.Although experienced physicians can identify the lesions in the image with professional knowledge,the presence of speckle noise affects the physician's judgment of the severity of the lesion.Study on ultrasonic image denoising based on variational partial differential equations,which can promote the understanding of image,but also can provide help for the diagnosis and treatment,to meet the people want a more intuitive understanding of the urgent needs of the severity of the disease,to build a bridge for communication between doctors and patients,reduce patient anxiety.This paper uses variational partial differential equation to solve the speckle noise removal in B ultrasound medical images,the main research contents and creations are as follows:(1)Combined with frame constraints on the solution of the ultrasonic image denoising model by Jin et al.(2011)and Split Bregman algorithm,a fast numerical algorithm for the model is given.The results of simulation show that the algorithm is feasible and effective,and it can greatly reduce the computation time in the removal of speckle noise.(2)A total variation model is proposed to maintain the detail information of B ultrasound images,and the existence and uniqueness of the solution is proved.Based on the Split Bregman algorithm,a fast algorithm is given.The experimental results show that the new method can preserve the details of the image while removing the speckle noise.(3)Combined with the two order model of B ultrasound images denoising and four order variational model,introducing control function,and proposing a new adaptive model to remove multiplicative noise in ultrasound images,numerical experiments show that the model can reduce the staircase effect and preserve edge information of the image.