Dissertation > Excellent graduate degree dissertation topics show

Image Denoising and Enhancement based on Curvelet Transform

Author: GuXiuPing
Tutor: ChenZaiPing；ZhangJingLei
School: Tianjin University of Technology
Course: Pattern Recognition and Intelligent Systems
Keywords: Curvelet Transform Image Denoising Image Enhancement Multi-scale Analysis
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 137
Quote: 0
Read: Download Dissertation

Abstract


Wavelet analysis has achieved great development in signal processing, which profits from its time-frequency analysis. But wavelet analysis can only deal with the situation where points are singularity, it can’t express features of an image along the edge effectively. Curvelet transform provides a new multi-resolution analysis method with various features, which represents edges better than wavelets and is therefore well suited for Multi-scale edge. Curvelet transform has high anisotropy, better description of edges and detail information of images. It has great potential for image processing.In this paper, firstly the background and significance in image denoising and enhanceme-nt are addressed. Then the classical methods and achievements are overviewed and summariz-ed. The basic theory of Curvelet transform, which belongs to the Multi-scale Gemetrical Anal-ysis (MGA), is discussed in depth. Two classical fast discrete Curvelet transform algorithms are analyzed. Based on these previous researches, the new algorithm of image denoising and enhancement are developed.In image denoising, a variety of image denoising algorithms are studied. According to the drawback of soft thresholding and hard thresholding denoising algorithms, a new method that compromise soft and hard thresholding approach is proposed. The Curvelets coefficients in different subbands are filtered with adaptive thresholds. We used the Peak Signal to Noise Ratio(PSNR) and subjective visual quality to estimate the image denoising quality. Compared with hard threshold denoising and soft threshold denoising based on wavelets domain, and other threshold regulation algorithms based on Curvelets transform domain, the proposed method can preserve more edges and texture of the reconstruction images. Experimental results show that our algorithm improves PSNR for 1.5dB than traditional method.In fog-degraded color images enhancement, firstly two classical enhancement algorithms—histogram equalization and multi-scale Retinex are reviewed. Then a new enhancement method based on vanishing point detection and second generation Curvelet transform is proposed.The experimental results show that the overall image enhancement is achieved and the image contrast is improved, which is better than the prior two methods.

Related Dissertations

  1. Research of Images Enhancing Algorithms on Fog or Backlighting Conditions and Implementation with Hardwares,TP391.41
  2. Research on Methods of Medical Ultrasound Image Denoising,TP391.41
  3. Application Research of Digital Image Processing on Container Inspection,TP274.4
  4. Technology of Blood Vessel Diameter Measurement Automatic Based on Digital Image Processing,R310
  5. Image Fusion Algorithms Based on Multi-scale Analysis,TP391.41
  6. Research on Contourlet Transform and Its Application on Image Processing,TP391.41
  7. Research on Pressure Sensed Fingerprint Identification System and Algorithem,TP391.41
  8. Wavelets-based Signal Sparse Representation and Its Application in Image Denoising,TP391.41
  9. Research of Image Enhancing of Vein Based on Fuzzy Theory,TP391.41
  10. Noise Analysis and Image Processing of CMOS Image Sensor,TP212
  11. Research on Pavement Crack Detection Algorithom Based on Digital Image Processing,TP391.41
  12. Research on Pavement Crack Detection Algorithm,TP274
  13. Based on kernel regression and non- local method of image denoising,TP391.41
  14. Action surface EMG nonlinear characteristic of,TH772
  15. Model based on visual characteristics and image enhancement algorithms and performance analysis,TP391.41
  16. Tetrolet sparse regularization learning with the sample image super-resolution algorithm,TP391.41
  17. PDE-based image denoising and enhancement of,TP391.41
  18. Infrared image enhancement and color lookup table -based technology research,TP391.41
  19. Multiscale information fusion algorithm,TP202
  20. Denoising problem solving variable weight fixed point algorithm,O177.91
  21. Research on the Pre-processing Methods for Vehicle Target Classification in SAR Images,TN957.52

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