本文選題：圖像復(fù)原　切入點：傅里葉變換　出處：《吉林大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著時代的不斷發(fā)展,人們對圖像質(zhì)量的要求越來越高.圖像復(fù)原技術(shù)作為改善圖像質(zhì)量的方法之一,已成為國內(nèi)外研究的熱點問題.圖像的復(fù)原技術(shù)廣泛地應(yīng)用于天文、醫(yī)學(xué)、軍事、航空等領(lǐng)域,具有重大的現(xiàn)實意義.本文以傳統(tǒng)圖像復(fù)原算法為基礎(chǔ),探討了幾種經(jīng)典的復(fù)原方法,提出了一種新的圖像復(fù)原算法,復(fù)原圖像的質(zhì)量在數(shù)值實驗和視覺效果上都有所提高.緒論中首先介紹了本文選題的研究意義,回顧了圖像復(fù)原技術(shù)的研究背景、國內(nèi)外的研究現(xiàn)狀以及圖像復(fù)原技術(shù)的最新進(jìn)展,緊接著給出了本文的整體框架.第二章介紹了圖像復(fù)原的基礎(chǔ)知識、圖像退化模型、退化函數(shù)估計、噪聲類型以及評價指標(biāo).第三章介紹了幾種經(jīng)典的圖像復(fù)原方法,比如維納濾波、逆濾波、約束最小二乘法等.第四章是本文的核心內(nèi)容,結(jié)合二對角變換以及尋找相似塊的方法,提出一種新的圖像復(fù)原方法.通過Householder變換二對角,使信號與噪聲有效分離.第五章對本文提出的算法進(jìn)行實驗仿真.通過實驗可以看出,本文提出的復(fù)原方法具有良好的效果,不僅保留了圖像的主要特征信息,并且在一定程度上抑制了噪聲.本文的最后一章對提出的復(fù)原算法進(jìn)行評價總結(jié)并對未來研究方向和重點做了進(jìn)一步的展望.
[Abstract]:With the development of the times, the demand for image quality is becoming higher and higher. As one of the methods to improve image quality, image restoration technology has become a hot issue in the research of domestic and foreign. Image restoration technology is widely used in astronomy. Medical, military, aviation and other fields have great practical significance. Based on the traditional image restoration algorithm, this paper discusses several classical restoration methods, and proposes a new image restoration algorithm. The quality of restored images has been improved in both numerical experiments and visual effects. In the introduction, the significance of this topic is introduced, and the research background of image restoration technology is reviewed. The research status at home and abroad and the latest development of image restoration technology are presented. Then the whole frame of this paper is given. In chapter 2, the basic knowledge of image restoration, image degradation model, degradation function estimation, and so on are introduced. Chapter 3 introduces several classical image restoration methods, such as Wiener filter, inverse filter, constrained least square method and so on. Chapter 4th is the core of this paper. A new image restoration method is proposed by combining the bidiagonal transformation and the method of finding similar blocks. In chapter 5th, the algorithm proposed in this paper is simulated. The experiments show that the restoration method proposed in this paper has a good effect, and not only retains the main feature information of the image, but also improves the performance of the image. In the last chapter of this paper, the proposed restoration algorithm is evaluated and summarized, and the future research directions and emphases are prospected.
【學(xué)位授予單位】：吉林大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP391.41

【參考文獻(xiàn)】

相關(guān)碩士學(xué)位論文 前1條

1 吳昊;基于曲波變換的離焦虹膜圖像復(fù)原算法的研究[D];吉林大學(xué);2013年

，

本文編號：1598782