Dissertation > Excellent graduate degree dissertation topics show

Research on Several Image Segmentation Methods Based on Fuzzy Entropy Theory

Author: ZhaoFeng
Tutor: FanJiuLun
School: Xi'an Institute of Posts and Telecommunications
Course: Signal and Information Processing
Keywords: Image Segmentation Fuzzy Entropy Generalized Fuzzy Entropy Image quality evaluation criteria
CLC: TP391.41
Type: Master's thesis
Year: 2007
Downloads: 478
Quote: 5
Read: Download Dissertation


Segmentation is the basic image processing and computer vision technology is an important part of most of the image analysis and visual system, is also a successful image analysis, to understand and describe the key steps. Fully consider the randomness of the image signal has several methods of image segmentation based on the probability entropy threshold selection, but ignore the blur contained in the image, the segmentation result is not satisfactory in many cases. Therefore, the problem of image segmentation is a basic and important to study the use of the fuzzy entropy resolved its internal necessity and rationality. Image segmentation using fuzzy entropy, fuzzy membership take split is 0.5, but when the images of uneven illumination, the limitations of the above assumptions is revealed. In this paper, based on generalized fuzzy entropy image segmentation method is taken as the fuzzy membership m (0 lt; m lt; 1) segmentation, which increase the opportunity to choose better segmentation results can be obtained more than the traditional fuzzy entropy segmentation method good segmentation results. This thesis research are summarized as follows: (1) for the Cauchy-type membership function determination algorithm more reasonable fuzzy domain, making the final threshold is closer to the lowest valley of the histogram; S-type membership function is given S-type two-dimensional membership function of a structure, the S-type two-dimensional membership function with respect to the existing S-type two-dimensional membership function has obvious geometric meaning, to be able to bring more ideal segmentation. (2) a combination of C-means clustering algorithm and fuzzy entropy image segmentation method, a combination of two-dimensional Otsu method and fuzzy entropy image segmentation method, a combination of two-dimensional entropy method and fuzzy entropy image segmentation method. These combined methods are those commonly used method of image segmentation initial image segmentation using fuzzy entropy reclassified preliminary split the wrong classification. For noisy images, this combined method will be superior to those commonly used in image segmentation method segmentation results. (3) gives the strict definition of generalized fuzzy entropy, constructed several generalized form of fuzzy entropy, and successfully applied to image segmentation, and achieved good segmentation. Finally, the use of image quality evaluation criteria to determine the definition of generalized fuzzy entropy of the parameter m (0 LT; m LT; 1) of the method, the experimental results show that, using this method to determine the m and to rely on visual observation of the results obtained is substantially uniform 's. The paper work was funded by the National Natural Science Foundation of China (No: 60572133).

Related Dissertations

  1. Research on Image Recognition Algorithm in the Forest Fire Prevention System,TP391.41
  2. Based on statistical methods of magnetic resonance imaging of the human brain image segmentation and three-dimensional analysis of the data,R445.2
  3. Road extraction algorithm based on region segmentation of remote sensing image,TP751
  4. Research for Infrared Image Target Identification and Tracking Technology,TP391.41
  5. The Study of Human Face Detection Method Based on OpenCV,TP391.41
  6. Research on Pavement Crack Detection Algorithom Based on Digital Image Processing,TP391.41
  7. Finger vein recognition technology,TP391.41
  8. Based on Information Fusion roads and obstacle detection method,TP242
  9. Solder paste printing quality based on image processing techniques to detect,TP391.41
  10. Research of Digital Image Segmentation,TP391.41
  11. Research of Fingerprint Preprocessing and Recognition Algorithm,TP391.41
  12. Markov random field DS evidence theory of the human brain image segmentation,TP391.41
  13. Image Processing and Automatic Recognition for Weld in X-ray Non-Destructive Testing,TP391.41
  14. Silk-Screen Quality Surveillance System Base on Image Processing Technology,TP391.41
  15. Research on the Prompt Steel Bar Separation System Based on Multi-view,TP391.41
  16. Application Research of Digital Image Processing on Container Inspection,TP274.4
  17. Texture Image Segmentation Based on Stationary Wavelet-Based Nonsubsampled Contourlet,TP391.41
  18. Data Reduction Algorithm Based on Information Entropy,TP18
  19. Temperature Recognition System of Thermal-paint Color Images Based on Color Image Processing Technology,TP391.41
  20. Chan-Vese active contour model of,TP391.41
  21. The Detection Study of Traffic Flow and Vehicle Queue Based on Video,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