Dissertation > Excellent graduate degree dissertation topics show

Research on Wavelet Compression Based on the Multiscale Geometric Analysis

Author: ZhouLinChuan
Tutor: ZhaoHui
School: Shandong University
Course: Signal and Information Processing
Keywords: Image Coding wavelet transform MGA smoothness
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 52
Quote: 0
Read: Download Dissertation

Abstract


In recent decades, with the development of computer technology, the higher demand is put forward by people in terms of the quality and volume of the digital image, therefore, the amount of image data increasing quickly, which brings big problems for data storage and transport. It’s necessary to compress the image to meet the needs of various applications. Lots of useful information contained in the image details, such as textures and contours. Different types of images contain different details, how to retain the useful information has become a hotspot of the current study. This research is aimed at the problem.Wavelet transform achieves great success in the image compression, but due to limitations of its theory, the ability of impress the details of high dimensional space is limited. This article introduces the basic theory of wavelet transform in details, and analyzes the characteristics of wavelet transform deeply. On the basis of wavelet transform, multiscale geometric analysis (MGA) theory achieves great development, and it has become one hot direction of the digital image compression research. MGA has the characteristics of anisotropy, and it can approach a high dimensional function containing line singularity effectively. This novel introduces the effective MGA tool——Contourlet transform. In a sense, Contourlet transform is a real 2D image sparse representation. It can impress the singularities of image effectively, and it performs well in image processing. Contourlet transform adopted double filter bank structure:first, using Laplace pyramid transform to do image multi-resolution analysis; second, using direction filter bank to do direction decomposition. LP decomposition is a redundant decomposition, so Contourlet transform does not suitable for image compression. To overcome this shortcoming, a new kind of MGA tool WBCT was proposed. WBCT combines wavelet transform and direction filter together realizes a no redundant transformation. Then, this article introduces two classic embedded coding methods:EZW and SPIHT.The WBCT-based SPIHT compression coding method restores the image directional character much better. It performs good to the image with more details, but for the image with little details, multiple directions decomposition will reduce the effect of decoded image. Therefore, this article puts forward the concept of smoothness, and the direction decomposition of wavelet sub-bands depend on smooth. According to the structure of coefficients from the algorithm proposed in this article, SPIHT algorithm was improved. Experiments show that the proposed algorithm can made different degree of direction decomposition to different types of images, and the decoded image is better than that form wavelet transform and WBCT.

Related Dissertations

  1. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  2. Research on Identification System of Cashmere and Wool Fiber,TS101.921
  3. Characteristics of sensory stimulation evoked,R318.0
  4. Network transmission ROI image coding algorithm,TN919.81
  5. Based on Contourlet Transform Digital Watermarking Method,TP309.7
  6. Image Fusion Algorithms Based on Multi-scale Analysis,TP391.41
  7. Feature Extraction Technologies Research and Implementation of 3D Models Based on Wavelet Transform,TP391.41
  8. Combined DWT Dynamic Data Reconciliation Research and Application,TP274
  9. The Compression and Application of the Brain CT Image,TP391.41
  10. Inspection and Recognition on a Class of Tool Processing Trace,TP391.41
  11. Building geometry and discrete spline modeling technology research,TU17
  12. CUDA-based video fire detection system,TP391.41
  13. Helicopter main gearbox fault diagnosis and failure prediction technology research,V267
  14. Ultra-precision double-sided polishing process stability studies,TG580.692
  15. The Methods and Applications of Constructing High-precision Quasi-interpolation Based on Bernstein Quasi-interpolation Scheme,O174.41
  16. Construction of Symmetric Complex Wavelets and Tight Framelets with Dilation 3,O174.2
  17. Hyperspectral Unmixing Based on Nonnegative Matrix Factorization,TP751.1
  18. Research and Development of System for Reconstruction and Optimization of 3D Mesh Model,TP391.41
  19. The Study on the Several Kinds of Convexity and Smoothness in Locally Space,O177.2
  20. B-valued martingale sequence nature and Martingale methods in financial markets,F830.9
  21. ICA based on redundant wavelet and image watermarking method,TP309.7

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