【正文】 回首這一段時期的緊張工作，不但令人難忘，而且深深感受受益良多，從中學(xué)到了許多寶貴的東西，深入研究了小波理論以及它在圖像處理方面的應(yīng)用，為今后的研究工作打下了良好的理論基礎(chǔ)。 ,只有當(dāng)閾值 選擇合適 ,才能使質(zhì)量明顯好轉(zhuǎn) ,信噪比有較大的提高。 figure,imshow(M2)。,M,39。den39。,4),J)/255。salt amp。對于紋理較多的圖像而言，由于 B2樣條雙正交小波的線性相位特性，對于圖像的細(xì)節(jié)部分的處理能較正交小波具有更小的失真，因此對于圖像去噪的結(jié)果也有一定的影響。 基于以上理由，在兩種小波去噪方式對比中，對正交小波采取硬閾值，而對雙正交小波在噪聲水平較低時采取硬閾值，而在噪聲水平較高時采取軟閾值。圖像的特點等具體情況有關(guān)，下面我們就小波基的正交性與線性相位性作一些簡單的討論。目前使用的閾值可以分成全局閾值和局部適應(yīng)閾值兩類。 figure,imshow(K2)。) （ a）原始圖像 （ b）加噪圖像 第 3章 （ c） 3*3中值濾波后的圖像 （ d） 9*9 中值濾波后的圖像 圖 3－ 2 調(diào)入 medfilt2 實現(xiàn)中值濾波 由圖 3－ 2 可以看出實踐表明 ， 9*9 中值濾波后的圖像去噪后的效果明顯比 3*3 中值濾波后的好，這說明，選擇適當(dāng)大小的中值濾波窗口可以在最大限度的保持圖像精度的基礎(chǔ)上去除圖像噪聲。,)。% 進(jìn)行 7*7均值濾波 figure,imshow(K2)。39。若尺度函數(shù)可分離，即：φ (x1，x 2)=φ (x1)*φ (x2)。通常情況下，基本小波ψ (t)以原點為中心，因此ψ a, b(t)是基本小波ψ (t)以 t=b為中心進(jìn)行伸縮得到。 。對有些處理過程來說，噪聲往往會產(chǎn)生某種局部二義性 (local ambiguities)。比較有影響的方法有： Eero moncelli 和 提出的基于最大后驗概率的貝葉斯估計準(zhǔn)則確定小波閾值的方法。 關(guān)鍵詞： 圖像去噪，均值濾波，中值濾波，維納濾波，小波變換，閾值 ABSTRACT In the formation, transmission and recording process of digital image, because imaging system, transmission media and recording devices are often imperfect images obtained are polluted by various noises. In pattern recognition, puter vision, image analysis and video coding and other areas, noise image preprocessing is extremely important, and whether its effect is good or bad will have a direct impact on the followup to the quality and analysis is a novel research field in the has a perfect performance on both local time and frequency domain,so it is an effective analytical image processing by the aid of wavelet analysis is one of the hot issues of wavelet research and a brief description of the traditional methods of image denoising,this paper makes a deep research on image denoising. In the paper,it introduces the status of image denoising, their developing trend and also the reasons and significance of imaged denoising. In the second chapter, it introduces the basic principles the image denoising and concepts of the image denoising ,which include classification of noise and its impact on image ,the evaluation of image denoising results and wavelet transform theory. Chapter 3 describes mon methods of the image denoising ,including the realization of Matlab code and analysis of the mean filter, filtering and Wiener filter. Chapter 4 of this essay is the most important part of the focuses on image denoising based on wavelet, In summing up the past experience of choosing thresholds,it introduces a new threshold estimation methodprobe method and the analysis of wavelet Selection based on the image denoising, and proves its use results. The final chapter tells a prehensive parative analysis of the mean filter, filtering, the Wiener filter and image denoising based on wavelet domain. The result is that the result of the image denoising based on wavelet domain was better than several other. This paper is divided into five chapters. The first chapter is the introduction. Chapter II introduces the basic principles and concepts of image denoising. Chapter III introduces mon methods and analysis of the image denoising. the fourth chapter introduces image denoising based on the wavelet domain and analysis. The final chapter introduces the prehensive parative analysis and conclusions. Keywords： emage denoising, filtering, median filtering, wiener filtering, wavelet transform , thresholding 摘要 ................................................................................................ 錯誤 !未定義書簽。摘要 數(shù)字圖像在其形成、傳輸和記錄過程中，由于成像系統(tǒng)、傳輸介質(zhì)和記錄設(shè)備的不完善往往使得獲取的圖像受到多種噪聲的污染。 ABSTRACT ................................................................................................................... 2 第 1 章 緒論 .................................................................................................................. 4 圖像去噪的現(xiàn)狀及發(fā)展趨勢 ......................................................................... 4 圖像去噪的研究現(xiàn)狀 ........................................................................ 4 基于小波域的圖像去噪的發(fā)展趨勢 ................................................... 5 研究圖像去噪的理由與意義 ......................................................................... 5 第 2 章 基本原理與概念 .............................................................................................. 6 圖像噪聲 ......................................................................................................... 6 去噪效果評價 ................................................................................................. 6 小波變換 ......................................................................................................... 8 連續(xù)小波 變換 ....................................................................................... 8 離散小波變換 .................................................................................... 9 第 3 章 圖像去噪的常用方法及分析 ...................................................................... 11 均值濾波 ....................................................................................................... 11 均值濾波的原理 ................................................................................ 11 中值濾波 ....................................................................................................... 12 中值濾波的原理 ................................................................................ 12 基于 matlab 中值濾波去噪方法的代碼實現(xiàn)及分析 ..................... 12 維納濾波 ....................................................................................................... 13 維納濾波的原理 ................................................................................ 13 基于 matlab 維納濾波去噪方法的代碼實現(xiàn)及分析 ........................ 14 第 4 章 基于小波域圖像去噪方法及分析 ................................................................ 15 圖像小波去噪原理 ...................