Research On Ultrasound Image Denoising And Segmentation Methods Based On Mathematical Morphology

Ultrasound imaging techniques in underwater exploration, medical diagnostics,industrial non-destructive testing and other areas are widely used.Ultrasound imagespeckle noise inherent in not only reduces the image quality, but also seriouslyaffected the image of manual interpretation and automatic identification, makingnoise and ultrasound image segmentation has become an important topic in the fieldof image processing.Mathematical morphology is one of a wide range of image processing methodsin recent years, the rise of the method has been widely used in image filtering, imagesegmentation, edge detection and other areas of basic knowledge, its comprehensivegeographic topology, mathematical set theory and other disciplines.In this paper, anultrasound image speckle noise, and the noise reduction process in the image willappear inaccuracy issues, studies based on mathematical morphology ultrasoundimage noise reduction algorithms.First, we design the opening and closing of the composite filter for filtering theultrasound image, the filters of different sizes, different shapes of structural elementswere filtered, and then re-integration of the filtered images, and ultimately get theultrasound image filtering based on the definition of image entropy images.Secondly, this paper an ultrasound image multiscale morphological gradientimage fusion method based on image entropy defined by the combination ofmethods and mark a watershed segmentation, the formation of a multi-scalemorphological gradient and ultrasound image segmentation based on watershedmark.The multi-scale gradient image using image entropy method uses the definitionwas better able to retain the characteristics of the image edge details, mark awatershed segmentation and ensure the accuracy of segmentation.In this paper, the ultrasound image noise reduction and segmentation methodbased on mathematical morphology theory carried out research, has achieved someresults, but the place is still insufficient, pending further study and improvement.