Dissertation > Excellent graduate degree dissertation topics show

Markov random field DS evidence theory of the human brain image segmentation

Author: GuoBin
Tutor: GuanYiHong
School: Kunming University of Science and Technology
Course: Physical Electronics
Keywords: Medical Image Segmentation Markov random field Fuzzy Clustering Fuzzy C- means algorithm D - S evidence theory Histogram Parameter estimation
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 105
Quote: 0
Read: Download Dissertation

Abstract


With the progress of the health care system, medical imaging has become an emerging discipline, medical image segmentation in medical image analysis plays an increasingly important role, is widely used in various fields of medical research. The basis of medical image segmentation is essentially the type of distribution of the tagging process to enter a variety of medical image voxel labeled image segmentation results, it is organizations measure the three-dimensional reconstruction, image registration, etc. clinical diagnosis has a certain role of adjuvant therapy. In medical imaging research, for the study of the human brain, has become the focus, because the brain is a human life and activity of the central nervous system, but due to the complexity of the organizational structure of the brain and irregularities, as well as the imaging process in the magnetic field uneven, etc. affect brain magnetic resonance images essentially blur these characteristics for magnetic resonance images of the brain, usually cited 入马尔科夫 image segmentation with the airport and fuzzy clustering theory and the Dempster-Shafer theory of evidence. This article is based on these three theories of brain MR image segmentation, this article reads as follows: 1. Research status and development trend of domestic and foreign medical image segmentation analysis of medical image segmentation difficulty and significance. Commonly used in medical image segmentation methods are introduced, magnetic resonance images of the brain, using image segmentation algorithm its split experiment, the result of the comparison. Brief introduction on the basis of medical imaging, especially computed tomography imaging principle and the magnetic resonance imaging principle, and gives the imaging characteristics and clinical application. 2 in-depth study of a Markov random field and Gibbs random brain magnetic resonance image segmentation. Markov in magnetic resonance images of the brain based on Markov random field theory based on magnetic resonance images of the brain spatial context information, experience in the energy field, two-dimensional histogram method and then combined with fuzzy clustering on the brain MRI image segmentation, extract the white matter and gray matter. Finally, discussion of the inadequacies of the segmentation results, the analysis method. 3. Markov random field to take into account the pixel neighborhood relations, the use of the spatial characteristics of the image, so the use of the split can get better segmentation results. The fuzzy clustering two-dimensional histogram method due to clustering theory and field pixels binding can get ideal segmentation results. Two methods classified dispute pixels within the fuzzy region segmentation with a magnetic resonance images of the brain when there will be a different division, the Dempster-Shafer theory of evidence to integrate information from different sources image, so this two ways to split the results label extraction as well as redundant image extraction, and then use the Dempster-Shafer theory of evidence to get the final results of the image of the original image and the label field and redundant image in accordance with the Dempster-Shafer fusion segmentation integrated decision-making rules, to resolve the dispute pixel classification problem. 4. Markov random field image segmentation applications biggest difficulty is how to select reasonable parameters more reasonable and accurate image modeling and segmentation, parameters including the form of the potential function is selected, the direction of the selected function β likelihood optimization function, so in this article to give some definition of the parameters of the method, and proposed a method of \spatial relationships between adjacent pixels and local interactions, and the likelihood function optimized for the likelihood function in the fuzzy sense, the introduction of membership in order to achieve a Markov random field parameters estimated, making the segmentation results between categories clear boundaries, more in line with the actual situation.

Related Dissertations

  1. Research on Direct Sequence Spread Spectrum Signal Detection and Parameter Estimation,TN914.42
  2. Modulation Pattern Recognition and Parameter Estimation for Ground Wave Radiation Sources,TN957.51
  3. Research of High Speed Image Pre-processing System Based on FPGA,TP391.41
  4. Research on Joint Target Detection for Dual-Sensor Image and System Implementation,TP391.41
  5. Research and Implementation on Content-Based Clothing Image Retrieval,TP391.41
  6. Research on Predictive Control and Simulation of pH Value Based on the Confection Process of Sodium Nitrate,TP273
  7. Based on statistical methods of magnetic resonance imaging of the human brain image segmentation and three-dimensional analysis of the data,R445.2
  8. The Research and Application of Software Reliability Test for User Right Management System,TP311.53
  9. Evolutionary Clustering Algorithm and Its Application,TP311.13
  10. Data Analysis of Multiple Repeated Measures of Children’s Behavior,O212.1
  11. Research of Color Correction Method for Multiview Video,TP391.41
  12. Study on Power System Voltage and Reactive Power Control Method,TM761.1
  13. HMM-based Connection between the Social Network Analysis,F49
  14. ESPRIT Parameter Estimation Algorithm in Wideband Channel System,TN925
  15. Cognitive radio spectrum detection optimization algorithm and MAC protocol design,TN925
  16. Based on information fusion safety assessment of genetically modified foods,TS201.6
  17. Online Banking Service Quality Evaluation Model,F830.4;F224
  18. Feature points based spatial information distribution histogram matching method,TP391.41
  19. MIDI-based instrument control systems and automatic identification method notes,TN912.34
  20. Study on Terrain Classification Based on MRF Model and Statistical Modeling of SAR Imagery,TN957.52
  21. Parameter Estimation of Nuclear Magnetic Resonance Spectroscopy Based on Reversible Jump MCMC,O482.532

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