Non-local mean filter method for three-dimensional ultrasonic images

Abstract

The invention discloses a filter method for three-dimensional ultrasonic images, and belongs to the field of three-dimensional image processing. The filter method includes loading a to-be-filtered three-dimensional ultrasonic image, splitting the to-be-filtered target three-dimensional ultrasonic image into reference blocks, and specifying a search area for each reference block; using all image blocks with the sizes identical to the size of each reference block in the corresponding search area as similar blocks and computing a similarity degree of each similar block and the corresponding reference block; using a weighted mean value of the corresponding similar blocks as a filter result of each reference block and a weight of the corresponding weighted mean value as a similarity degree of each reference block; integrating the filter results of the reference blocks to obtain a final filter result of the three-dimensional ultrasonic image. Compared with the prior art, the filter method has the advantages that the three-dimensional image can be directly filtered without being split into two-dimensional image frames to be filtered, and gray information among frames can be utilized, so that a filter effect can be obviously improved, and detail information of the image can be effectively preserved; the filter method is applicable to accelerating a GPU (graphics processing unit), so that the filter time can be greatly shortened.

Description

Technical field [0001] The invention belongs to the field of three-dimensional image processing, and more specifically, relates to a non-local mean filtering method for three-dimensional ultrasound images. Background technique [0002] Three-dimensional ultrasound imaging has achieved research results in many aspects including basic theories, imaging techniques, clinical experiments, etc., and it has received more and more attention in clinical diagnosis. Three-dimensional ultrasound imaging technology can provide three-dimensional structure images of tissues and organs. Compared with traditional two-dimensional ultrasound, the image display is intuitive, which facilitates accurate measurement of medical diagnostic parameters. However, due to the coherent characteristics of ultrasound imaging, speckle noise exists in the image, and during the process of acquiring, converting, and transmitting three-dimensional ultrasound images, it is susceptible to interference from imaging equi...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above defects or improvement needs of the prior art, the present invention provides a non-local mean filtering method for 3D ultrasound images. grayscale information, thus solving the technical problems of poor filtering effect and loss of detail information in current 3D image filtering methods

Images