Medical ultrasonic image denoising method based on wavelet transform and trilateral filter

Abstract

The invention provides a medical ultrasonic image denoising method based on wavelet transform and a trilateral filter. The medical ultrasonic image denoising method comprises the following steps that step 1) a medical ultrasonic image model is established; and step 2) wavelet decomposition is performed on an image obtained in the step 1 after logarithmic transform so that four frequency domains (LL1, LH1, HL1 and HH1) are obtained. Wavelet decomposition is continuously performed on the low frequency domain LL1 so that four frequency domains (LL2, LH2, HL2 and HH2) are obtained again; then the step is repeated until the maximum number J of layers of decomposition; step 3) threshold method contraction processing is performed on the wavelet coefficient of the high frequency part (LHj, HLj and HHj, j=1,2,...,J) of each layer; step 4) filtering processing is performed on the wavelet coefficient of the low frequency part (LLJ) of the last layer by utilizing the trilateral filter; and step 5) inverse wavelet transform processing is performed so that a denoised medical ultrasonic image is obtained.

Description

technical field [0001] The invention is applied to the field of medical ultrasonic image denoising, and a wavelet transform-based denoising method suitable for medical ultrasonic images is designed. Background technique [0002] With the development of science and technology, in the field of medical imaging, imaging technologies such as ultrasound imaging, CT, and MRI have been applied in medical clinical diagnosis. Due to the rapid development of ultrasound imaging technology due to its advantages of high resolution, simple operation, instantaneity, etc., it has the characteristics of non-invasive, no radiation damage, fast and convenient, and has become a widely used and highly safe medical diagnostic technology. Especially for the examination of human body organs and tumor tissue, the use of ultrasound imaging technology is more important. [0003] Due to the limitation of ultrasound imaging mechanism, the existence of speckle noise seriously affects the quality of ultra...

AI Technical Summary

Problems solved by technology

Through experiments, although the complexity of the adaptive filtering method is low, it often blurs the details of the image, and the suppression effect on speckle noise is not very ideal.

The anisotropic diffusion denoising method has a strong denoising ability, but the result may be over-smoothed

The non-local denoising algorithm has an ideal suppression effect on speckle noise, but this kind of denoising method has high complexity and is not easy to meet the real-time requirements of medical ultrasound imaging system, so it is often used in the post-denoising processing of medical ultrasound images

Although the bilateral filter can keep the edge information very well and is recognized by everyone, but due to the high computational complexity of the bilateral filter and the phenomenon of "gradient distortion", it cannot meet the real-time requirements of the medical ultrasound imaging system.

Images