Research On Ultrasonic Image Denoising Method Based On Machine Learning

Ultrasound imaging technology has gradually become one of the most widely used diagnostic tools in medical clinics due to its advantages of low cost,high real-time performance,and no damage to the human body.However,due to factors such as the imaging principle of the ultrasound imaging system,the resulting ultrasound image is contaminated with speckle noise.Speckle noise blurs the detailed features of the image,thereby reducing the contrast and quality of the image to a certain extent.Handling is inconvenient.Therefore,studying how to reduce speckle noise in ultrasound images is of great practical significance for clinical diagnosis and image processing.By studying the distribution characteristics of speckle noise in the image and its forming factors,some denoising techniques in medicine can be used to effectively suppress and reduce speckle noise in the ultrasound image,so that the normal tissue parts,pathological parts and other parts of the human body in the image High contrast.In order to reduce speckle noise in ultrasound images,after consulting a large number of documents and materials,the research content of this article is determined.Firstly,this paper proposes an adaptive filter noise suppression method based on support vector machine.It uses mean filtering to maximize the smoothness of the noise area in the image,and can effectively retain the edge features in the original image,so as to achieve the purpose of distinguishing the noise area and the tissue area in the original image.Comparing the processing results of this method and the classic method through a certain evaluation index,although there is a certain improvement in signal-to-noise ratio and contrast,the improvement effect is not very good.By studying the rapid bilateral filtering method proposed by Shao et al.And the maturity of BP neural network,this paper once again proposes an adaptive bilateral filtering method based on BP neural network.This method uses the adjustable nonlinear function of BP neural network Linear characteristics,rapid prediction of the adaptive threshold of the bilateral filter,so as to achieve the adaptive distinction between the noise area and the tissue area,to achieve maximum smoothness of the ultrasound image while retaining its edge characteristics.Compare the experimental results with the traditional mean filtering;the classic P-M model and DPAD method,this method can suppress speckle noise in the ultrasound image with a faster processing speed,and retain the edge details in the image.