簡易檢索 / 詳目顯示

回結果列表
研究生： 陳亮亨
Chen, Liang-heng
論文名稱： 低解析度與曝光不足之影像合成演算法
Image Synthesis by Low Resolution and Underexposure Images
指導教授： 戴顯權
Tai, Shen-Chuan
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 電腦與通信工程研究所
Institute of Computer & Communication Engineering
論文出版年： 2007
畢業學年度： 95
語文別： 英文
論文頁數： 56
中文關鍵詞： 晃動 、影像合成 、短時間曝光 、小波
外文關鍵詞： underexposure, hand-shaking, wavelet, image synthesis
相關次數： 點閱：118  下載：1 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 本篇論文的主要研究，是將兩張取景相同的影像合成。這兩張影像皆為短時間曝光，而短時間曝光的優點之一就是可以避免手持相機所造成的晃動。由於曝光時間的不足，因此高解析度影像的整體對比較低，亮度也不足；另一張影像則因為解析度較低，在相同的曝光時間就可以有足夠的亮度。
    這兩張影像各有其資訊，大圖雖然在亮度與色彩上表現地不如小圖，但是卻有著小圖沒有的細節。只是單純地加強大圖的對比，也會同時突顯雜訊信號；如果直接將小圖利用內插演算法放大，也沒有辦法計算出遺失的細節部份。因此若能夠將這兩張圖的優點結合，就能夠產生一張在解析度與亮度皆有不錯表現的影像。但是實際上兩張影像之間拍攝的時間差有可能造成位移，以及相機在亮度不足的情況下所產生的雜訊問題，都是必須要解決的問題。
    小圖所含有的對比資訊是屬於低頻訊號，而大圖的細節資訊則是高頻訊號。本篇論文利用小波分解和多重解析度分析為基礎，提出了一個影像合成的演算法，能夠利用這兩張影像來避免拍攝時所造成的晃動。

    In this thesis, an algorithm is proposed to synthesize an image from two images of the same scene with short exposure time. One of the advantages is to avoid hand-shaking e ect of cameras. One image is high-resolution and
    has detail information but is underexposed; the other image is well-exposed therefore the resolution is much lower.
    Each image has its excellences, the HR (high-resolution) image has edge information, and the LR (low-resolution) image has precise luminance, chroma, and contrast. If the contrast enhancement algorithm is applied directly to
    the HR image, it also enhances the noise signals; up-sample the LR image by interpolation algorithms seems good, but the edges are blurred. In practical, the images taken at di erent time may causes motions, and noises arise
    because of underexposure.
    Contrast information is low-frequency signal and edge information belongs to high-frequency part. The proposed algorithm synthesizes images based on wavelets decomposition and multi-resolution analysis.

    Contents i List of Tables iii List of Figures iv 1 Introduction 1 2 Related Works 3 2.1 Contrast Enhancement . . . . . . . . . . . . . . . . . . . . . . 3 2.1.1 Histogram . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.1.2 Histogram Equalization . . . . . . . . . . . . . . . . . 3 2.1.3 Histogram Matching . . . . . . . . . . . . . . . . . . . 6 2.2 Resolution Enhancement . . . . . . . . . . . . . . . . . . . . . 7 2.2.1 Up-sampling by Interpolation . . . . . . . . . . . . . . 7 2.2.2 Super-Resolution . . . . . . . . . . . . . . . . . . . . . 9 2.3 Edge Detection . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.4 Discrete Wavelet Transform . . . . . . . . . . . . . . . . . . . 12 2.5 Scale-Invariant Feature Transform . . . . . . . . . . . . . . . . 15 2.6 De-noising Filters . . . . . . . . . . . . . . . . . . . . . . . . . 19 3 Proposed Algorithm 20 3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2 Contrast Enhancement . . . . . . . . . . . . . . . . . . . . . . 23 3.3 Motion Estimation . . . . . . . . . . . . . . . . . . . . . . . . 24 3.3.1 Feature Detection . . . . . . . . . . . . . . . . . . . . . 24 3.3.2 A ne Transformation . . . . . . . . . . . . . . . . . . 26 3.4 Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.4.1 Synthesis by Discrete Wavelet Transform . . . . . . . . 29 3.5 Noise Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.5.1 Edge-Preserved De-Noising by Nonlinear Curve Syn- thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4 Experiments 33 4.1 Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 4.2 Synthesized Images . . . . . . . . . . . . . . . . . . . . . . . . 33 4.2.1 Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.2.2 Room . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.2.3 Drum . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.2.4 Game . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.2.5 Rose . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 5 Conclusion and Future Works 51 5.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 5.2 Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Bibliography 54 Curriculum Vitae 56

    [1] Subhasis Chaudhuri, editor. Super-Resolution Imaging. Kluwer Aca-
    demic Publishers, 2002.
    [2] Rafael C. Gonzalez and Richard E. Woods. Digital Image Processing.
    Prentice Hall, 2002.
    [3] Bahadir K. Gunturk and Murat Gevrekci. High-resolution image recon-
    struction from multiple di erently exposed images. IEEE Signal Pro-
    cessing Letters, 13(4):197{200, April 2006.
    [4] Hsieh S. Hou and Harry C. Andrews. Cubic splines for image interpola-
    tion and digital ltering. IEEE Transactions on Acoustics, Speech, and
    Signal Processing, 26(6):508{517, December 1978.
    [5] M. Imme. A noise peak elimination lter. CVGIP: Graph. Models Image
    Process., 53(2):204{211, 1991.
    [6] Apogee Instruments Inc. Pixel binning.
    http://www.ccd.com/ccd103.html.
    [7] T.-L. Ji, M. K. Sundareshan, and H. Roehrig. Adaptive image contrast
    enhancement based on human visual properties. IEEE Transactions on
    Medical Imaging, 13(4):573{586, December 1994.
    [8] Robert G. Keys. Cubic convolution interpolation for digital image pro-
    cessing. IEEE Transactions on Acoustics, Speech, and Signal Processing,
    29:1153{1160, December 1981.
    [9] Yu-Mao Lin. Image stabilization system on a camera module with image
    composition. Master's thesis, National Taiwan University, 2006.
    [10] ImageMagick Studio LLC. Imagemagick.
    http://www.imagemagick.org/.
    [11] David G. Lowe. Distinctive image features from scale-invariant key-
    points. International Journal of Computer Vision, 60:91{110, November
    2004.
    [12] St ephane Mallat. A Wavelet Tour of Signal Processing. Academic Press,
    1999.
    [13] Don P. Mitchell and Arun N. Netravali. Reconstruction lters in com-
    puter graphics. In International Conference on Computer Graphics and
    Interactive Techniques, volume 22, pages 221{228, August 1988.
    [14] Khlifa Nawr es, Hamrouni Kamel, and Ellouze Noureddine. Image de-
    noising using wavelets: A powerful tool to overcome some limitations
    in nuclear imaging. In Information and Communication Technologies,
    2006. ICTTA '06. 2nd, volume 1, pages 1114{1118, April 2006.
    [15] Wei-Ting Sun. An e cienct and adaptive noise removal algorithm. Mas-
    ter's thesis, National Cheng Kung University, 2007.
    [16] Andrea Vedaldi. Sift++ { a lightweight c++ implementation of sift.
    http://vision.ucla.edu/ vedaldi/code/siftpp/siftpp.html.

    下載圖示
    2008-06-27公開
    QR CODE