An Adaptive Median Filtering Method for CT Image Denoising

Abstract

The invention discloses a CT image denoising method based on adaptive median filtering, belongs to the technical field of medical image processing, and is particularly suitable for CT image denoising of new coronary pneumonia. Select a square filter window with a size of n×n, compare the adaptive maximum and adaptive minimum of the gray value in the window with the current pixel gray value f(i,j), according to the first threshold T 0 Judging whether the current pixel point is a suspected noise point, if so, according to the second threshold T 1 Further accurately judge whether it is a noise point; if the current pixel point is not a suspected noise point or a noise point, traverse the next pixel point in the window; process the noise point through the center weighted median filter method; finally output the median filter to denoise After CT image. While maintaining image denoising, the image details are better protected; the problem of information deviation in the traditional entropy weight method is corrected by improving the entropy weight method, and the optimal value of the center weighted filter is determined by the contribution value of each evaluation index to the denoising effect. optimal weight to achieve the best denoising effect.

Description

Technical field [0001] The invention relates to the technical field of image signal processing, in particular to a CT image denoising method based on adaptive median filtering, which is suitable for the filtering processing of pneumonia CT images, especially for COVID-19 CT images, and belongs to the technical field of medical image processing. technical background [0002] COVID-19's lesions are mainly characterized by various ground-glass shadows or consolidation shadows, and CT images will be disturbed by impulse noise during transmission and acquisition. The changes of early lesions in COVID-19 are not obvious, the number of lesions is small, the scope of lesions is small, and the density is low. Impulse noise interference easily leads to missed diagnosis of early COVID-19 patients. [0003] According to the traditional CT image impulse noise denoising algorithm, the median filter is easily affected by the window size. When the window size is too small, the median filter is e...

AI Technical Summary

Problems solved by technology

In the early stage of the new crown, the changes of the lesions are not obvious, the number of lesions is small, the range of the lesions is small, and the density is low. Impulse noise interference is likely to cause the problem of missed diagnosis of early new crown patients

[0003] For the traditional CT image pulse noise denoising algorithm, the median filter is easily affected by the window size. When the window size is too small, the median filter is easily disturbed by the surrounding pixels, so that the gray value of the original signal point is replaced by noise. The gray value of the point makes the image denoising ability limited, and the noise cannot be removed correctly

When the window size is too large, the interference ability of noise points will be greatly reduced, which can greatly improve the image denoising ability, but will cause image details (such as edges, lines, corners, etc.) to be destroyed

Therefore, traditional median filtering is difficult to perform well in image denoising and preserving image details.

Images