透過您的圖書館登入
IP:18.222.144.244
  • 學位論文

利用超級像素分類技術估計數位影像之高斯雜訊

Gaussian Noise Estimation with Superpixel Classification in Digital Images

指導教授 : 張恆華

摘要


數位影像在擷取傳輸的過程中,會因為各種因素的干擾而產生雜訊,而有些因素則是無法避免的。這些影像中的雜訊,除了對影像視覺上的品質產生影響外,也會降低影像處理的精確度。因此許多影像處理的演算法,在處理前都會對影像做去雜訊的處理。其中,多數的去雜訊演算法會使用雜訊資訊,作為處理過程中的參數。除此之外,在影像處理及電腦視覺的領域中,許多演算法都會使用到雜訊程度的資訊。現今,大部分影像雜訊估計的方法,對於低雜訊的影像會有過度估計的現象。本論文之目標為提出一個新的方法以提升低雜訊影像估計的精確度。我們假設雜訊的模型為高斯雜訊,所提出的方法則是以其統計特性作為方法設計的基礎。首先,使用正規化剪切方法將影像的像素群聚成超級像素。接著,透過這些超級像素內的統計特性,判斷其是否與雜訊相似。從這些超級像素中找出具代表性的區塊,並以其資訊計算估計結果。我們使用不同類型的數位影像及柏克萊影像資料庫進行實驗,實驗結果顯示本研究所提出的雜訊估計方法,在低雜訊的影像能有更精確的估計結果,在雜訊增加時也能保有一定的精確準度。

並列摘要


Noise estimation is essential in a wide variety of digital image processing applications. It provides an adaptive mechanism for many restoration algorithms instead of using fixed values for the amount of noise. In this thesis, we propose a new statistical method based on the superpixel maps for estimating the variance of additive Gaussian noise in images. The proposed approach consists of three major phases: superpixel classification, local variance computation, and statistical determination. Experimental results suggest that the proposed method provides good estimation and is of potential in many image restoration applications that require automation.

參考文獻


[1] I. Pitas, Digital Image Processing Algorithms and Applications: Wiley, 2000.
[2] H. H. Chang, "Entropy-based trilateral filtering for noise removal in digital images," in Image and Signal Processing (CISP), 2010 3rd International Congress on, 2010, pp. 673-677.
[3] R. Garnett, T. Huegerich, C. Chui, and H. Wenjie, "A universal noise removal algorithm with an impulse detector," Image Processing, IEEE Transactions on, vol. 14, pp. 1747-1754, 2005.
[4] S. Baker and I. Matthews, "Lucas-Kanade 20 Years On: A Unifying Framework," Int. J. Comput. Vision, vol. 56, pp. 221-255, 2004.
[5] G. Calvagno, F. Fantozzi, R. Rinaldo, and A. Viareggio, "Model-based global and local motion estimation for videoconference sequences," Circuits and Systems for Video Technology, IEEE Transactions on, vol. 14, pp. 1156-1161, 2004.

延伸閱讀