Research And Implementation Of Ultrasound Image Denoising And Enhancement Algorithm

Medical ultrasound equipment has the advantages of low price,high real-time performance and no radiation.It has been widely used in human pathological examinations.With the advancement of medical technology,medical ultrasound image plays an increasingly important role in medical clinical diagnosis and treatment.But due to the limitations of its own imaging principle,ultrasound image has the disadvantages of low resolution and high image speckle noise density,which not only hinder the analysis of pathological features of patients by doctors but also affects the extraction of ultrasound image features by computer technology.However,traditional medical ultrasound equipment is bulky and has little mobility,making it difficult to perform its functions of some emergency situations.On the one hand,aiming at the speckle noise density in ultrasound images,an ultrasound image de-noising algorithm based on multi-directional statistical properties is proposed.The algorithm converts the multiplicative noise of the original ultrasound image into additive noise using logarithmic changes.Then the image is decomposed through multi-scale transformation to obtain multiple high-frequency sub-bands and one low-frequency sub-band,and the gradient domain guided filtering is used to process the low-frequency sub-bands to reduce noise and preserve image details.Sub-band,use the characteristics of sub-bands in different directions in the high-frequency signal to determine the appropriate threshold,and then optimize the shrink function to effectively reduce the noise in the high-frequency sub-band area on the premise of retaining the detail information,and obtain noise reduction through inverse multi-scale transformation.image.In the algorithm comparison test,the images with ideal speckle suppression effect collected by the ultrasound diagnostic system LOGIQ S8 XDclear 2.0+ of the hospital’s clinical ultrasound imaging department were used as the experimental data,and the noise variance was 0.03,0.04,0.05 on the dataset PSNR/SSIM They are respectively improved to32.3931/0.92015,31.8764/0.90490,and 31.24685/0.893075.From the comprehensive analysis of subjective visual effects and objective indicators,the de-noising effect of the improved algorithm is better than that of the compared algorithms.On the other hand,aiming at the problems of low definition and low contrast of ultrasound images,an improved contrast-limited histogram equalization ultrasound image enhancement algorithm based on energy distribution characteristics is proposed.The gray level of the image is limited by counting the energy change characteristics of the ultrasound image.The algorithm uses the limited contrast histogram equalization algorithm to pre-enhance the ultrasound images.The energy changes in the ultrasound images are then analyzed and normalized.The normalized values are introduced into the sigmoid equation.Finally,the calculation result is used as the weight of the pre-enhanced image to obtain the final enhanced image.Based on the data set of the hospital’s clinical ultrasound imaging department,the enhancement algorithm improves the information entropy and average gradient index values to 7.0105 and 6.9850,respectively.Improve the clarity and contrast of the image,with better visual effects.Based on the study of ultrasonic image de-noising algorithm and ultrasonic image enhancement algorithm,according to the characteristics of the algorithm and the actual situation of the equipment,this paper completes the compilation of ultrasonic image processing function code on the mobile platform,provides users with portable real-time ultrasonic image de-noising and enhancement processing results,solves the problem of large volume of ultrasonic testing equipment,improves the contrast and clarity of the image,and enhances the visual effect of ultrasonic image.