Dissertation > Excellent graduate degree dissertation topics show

Segmentation of Retinal Blood Vessels Based on Data-Driven Markov Chain Monte Carlo

Author: ZhouZuo
Tutor: ChenSongCan
School: Nanjing University of Aeronautics and Astronautics
Course: Applied Computer Technology
Keywords: Markov Monte Carlo Data-driven Vessel segmentation Curvelet transform Medical Image Processing
CLC: R318.0;TP391.41
Type: Master's thesis
Year: 2011
Downloads: 83
Quote: 0
Read: Download Dissertation

Abstract


Retinal blood vessels as the body non-invasive observation of the vital organs , which can reflect the different degrees of change in blood pressure, atherosclerosis and other cardiovascular diseases, symptoms, particularly when associated with vascular lesions occur in organs , retinal vessel diameter , curvature , etc. This change in characteristics to some extent, reflect the extent of lesions . Therefore, quantitative and qualitative analysis of retinal vascular automatically has a very important clinical value, and retinal vessel segmentation and extraction of blood vessels is the primary task . Most existing segmentation method although retinal vascular lesions of non- vascular image segmentation has good effect, but the effect of the lesion image segmentation is still not ideal , You Duiguang according uneven lesions and other sensitive . To this end , this paper presents a relatively robust vascular segmentation. This method is the first attempt to use computer vision Top-down and Bottom-up two kinds of hierarchical segmentation framework by combining the retinal image segmentation. Specifically, first, the retinal image of green color channel transform using Curvelet vessel enhancement . Then Bayesian statistical framework , using a reversible jump Markov Monte Carlo algorithm to search the parameter space , thus obtained does not depend on the initial segmentation approximate global optimal segmentation, while taking advantage of data-driven mean shift clustering algorithm and Canny edge detection operator to accelerate Markov chain dynamics . This article in the MATLAB environment, using standard STARE retinal image library four images , the experiment results show that the method is not only images but also for non-diseased lesion images are robust to segmentation , and through pattern recognition There is no free lunch theorems of the method is , in theory, the theoretical analysis .

Related Dissertations

  1. Image Fusion Algorithms Based on Multi-scale Analysis,TP391.41
  2. Fundus Image Segmentation Based on SVM and Template Matching,TP391.41
  3. SAFFRON based automated testing framework for QTP research,TP311.52
  4. Scrap copper smelting process control system design and implementation,TP273
  5. OCT fundus image layer segmentation and detection of OD Center,TP391.41
  6. Based on DCE-MRI of computer-aided diagnosis of benign and malignant breast lesions studied,TP391.7
  7. Automated Testing Framework System ATestPPMC Research and Implementation,TP311.53
  8. Application of Bayesian Method in Risk Based Inspection,X928.03
  9. Particle Filter Algorithm in Intelligent Transport System Research,TP29-AC
  10. Road Extraction Technolegy Based on Multi-Sensor Image Fusion,TP391.41
  11. Vascularity-Oriented Level Set Algorithm for Vessel Segmentation in Medical Image Processing,TP391.41
  12. The Study on Seismic Data Denoising and Interpolation with Curvelet Thresholding Iterative Method,P631.44
  13. A Bayesian Probabilistic Approach to Structural Parameter Identification and Application,O212
  14. The application of image processing technology in the study of neurological and psychiatric disorders,R749
  15. Research on the Piecewice Constant Level Set Method and the Mumford-Shah Model for Medical Image Segmentation,TP391.41
  16. Study on Outstanding Claims Reserve of R&C Based on Generalized Addictiv Mixed Model,F840
  17. Based on the second generation Curvelet transform image fusion research,TP751
  18. Corpus vocabulary teaching in high school experimental study,G633.41
  19. Automated software testing technology and applied research,TP311.52
  20. Based on B / S structure of the study and implementation of automated testing,TP311.52

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