Dissertation > Excellent graduate degree dissertation topics show

Research on Methods of Medical Ultrasound Image Denoising

Author: ChenZuoYi
Tutor: ShenYi
School: Harbin Institute of Technology
Course: Control Science and Engineering
Keywords: image denoising the ultrasound image the median filter mean filter the anisotropic diffusion
CLC: TP391.41
Type: Master's thesis
Year: 2008
Downloads: 300
Quote: 0
Read: Download Dissertation


Image denoising is widely used in the field of image preprocessing, and its main purpose is to improve signal-to-noise ratio and to highlight the expectation features for images. The medical ultrasound image denoising is an important field of medical ultrasound image processing, and also significant for general image processing. Depressing the speckles has always been one of the most key subjects of ultrasound imaging technology home and abroad. It is mainly desired that the useful image details for post-analysis and diagnosis can be preserved when the speckles are efficiently depressed.An improved median filter algorithm was proposed for the images highly corrupted with salt-and-pepper noise. All the pixels were divided into signal pixels and noise pixels using the Max-min noise detector, and the noise pixels were separated into three classes based on the local statistic information, which were low-density noises, moderate-density noises and high-density noises. Different filter algorithms were implemented to the noises of corresponding density level, and the weighted eight-neighborhood similarity function filter was implemented for the low-density noises due to its good detail-preservation. The proposed algorithm has better performance in the noise removal capability, adaptive capability and detail-preserving capability, especially being effective for the cases where the images are extremely highly corrupted.A weighted mean filter algorithm was proposed based on histogram and neighboring correlativity coefficients. The gray histogram was introduced as weight value of mean filter and it offered gray distribution information of original images. The correlativity coefficients were introduced due to its correlativity to distinguish the signals and noises. The proposed algorithm adaptively adjusts its filtering intensity according to its local signal feature, and powerfully reduces salt-and-pepper noise and preserves the image details.A denoising algorithm based on median-anisotropic diffusion was proposed for ultrasound images corrupted with speckles. Multi-directional median filtering was used to remove speckles and preserve edge details owing to its fine edge-preservation capability and the diffusion coefficient incorporating with unitary local variance and image gradient was introduced to improve local adaptation of diffusion modeling. The proposed algorithm has good performance in terms of speckle removal, edge-preservation and iteration speed.A non-subsampled contourlet transformation denoising algorithm was proposed based on context modeling. Fine characteristics of contourlet transformation were used such as multi-direction, high redundancy and translation-invariant. Spatially Adaptive context modeling can efficiently estimate and determine variance and threshold of the coefficient matrixes after the transformation and showed good local adaptation and accuracy. The soft-thresholding was employed on the detail subbands of different scales and directions to remove noises, and inverse non-subsampled contourlet transformation was used lastly. This proposed algorithm gains good visual effect and denoising effect.

Related Dissertations

  1. Research of High Speed Image Pre-processing System Based on FPGA,TP391.41
  2. Research on Contourlet Transform and Its Application on Image Processing,TP391.41
  3. Wavelets-based Signal Sparse Representation and Its Application in Image Denoising,TP391.41
  4. Research on Noise Reducation and Color Interpolation Algorithms Based on Bayer Pattern Images for Digital Camera,TP391.41
  5. Based on kernel regression and non- local method of image denoising,TP391.41
  6. Tetrolet sparse regularization learning with the sample image super-resolution algorithm,TP391.41
  7. PDE-based image denoising and enhancement of,TP391.41
  8. Denoising problem solving variable weight fixed point algorithm,O177.91
  9. Hypercomplex sparse representation and its applications,TP391.41
  10. Of speckle noise removal in medical ultrasound image,TP391.41
  11. Design of Image Acquisition and Processing System of Ultrasonic Endoscope Based on USB Isochronous Transfers,TH776.1
  12. Study on Image Processing and Imaging Technology of the Ultrasonic Phased Array Testing System,TH878.2
  13. Study on Segmentation of Breast Tumor in Ultrasonogram,TP391.41
  14. Research on ECG Real-time Detection Algorithm and Application,TH772.2
  15. Image Denoising and Enhancement based on Curvelet Transform,TP391.41
  16. Color Sort Method Research Based on SOPC Technology,TP391.41
  17. A RBF neural network - based image processing method,TP391.41
  18. Research on Image Steganalysis for LSB Matching,TP391.41
  19. Research on the Denoising and the Restoration of Frame Interpolation of X-band Nautical Radar Sea-surface Image,TP391.41
  20. Study on the Recognition and Classification System of Cervical Exfoliated Cells Based on Neural Network,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile