簡易檢索 / 詳目顯示

回結果列表
研究生： 梁靖汶
Jin-Wen Liang
論文名稱： 鄰近相關三角形 - 一個對於影像感測器中的壞像素偵測與還原之方法
Proximity Triangle - A Bad-Pixel Detection and Repair Scheme for Color Image Sensors
指導教授： 黃錫瑜
Shi-Yu Huang
口試委員:
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 電機工程學系
Department of Electrical Engineering
論文出版年： 2007
畢業學年度： 95
語文別： 英文
論文頁數： 51
中文關鍵詞： 影像還原 、壞像素 、影像感測器
外文關鍵詞： image repair, image filter, bad pixel, sensors
相關次數： 點閱：2  下載：0 
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

    • 隨著市場上對於影像能夠高解析的要求越高，在應用上，影像感測器的數目也就會越來越多。由於無論製成如何的精密，也不可能做到所有的影像感測器都能完美的達到要求。因此，這些不完美的影像感測器就會產生有瑕疵的影像。所以，一直以來，還原瑕疵的影像的方法一直不斷地被提出和創新。然而，過去有許多的還原方法都著重於處理過多的瑕疵，以致於也使得許多正確的資料也被誤判且改變。也因此，我們提供了一種方法，能夠在盡量不影響到正常像素的情況下，自動抓到壞的像素並且自動還原，使人眼看不出來還原後的影像有任何的瑕疵。
    不同於過去大多數常用的單色處理方式，我們提出的方法是利用色彩三元素(RGB)之間的關係，以達到只抓到壞像素的目標。經由這樣的關係，我們成功的避免了誤判正常像素的情況，並且能夠跟過去的方法一樣有效的抓到真正有壞的像素。
    在處理較少瑕疵的影像中，上述的方法可以做的非常的好，然而在處理較多瑕疵的影像中，上述的方法判斷壞像素的能力就明顯下降。為了作這方面的補強，我們也提供了另一個可以輔助上述色彩三元素相關的方式，可以使得判斷壞像素的能力大大的提升，且不至於誤判太多的正常像素。
    在本文的實驗結果顯示，我們的方式在越少壞像素的情況下會比別人的方法還好，這是因為我們的方法可以有效避免誤判正常像素。這也說明了許多過去的方式往往忽略誤判正常像素的問題。我們用Lena彩色的圖做測試，並且加入了少許壞像素的情況下，我們的方法比別人的方法好約2~15 dB。

    Image repaired is one of the important problems of image processing. In this paper, we propose bad-pixel detection and repair scheme for color image sensors based on a new proximity triangle concept. Unlike previous methods that mostly utilize the spatial correlation as identifying the faulty pixels for repair, we further exploit the chromatic correlations among the three fundamental colors (i.e., RGB). By doing so, we can avoid misclassification of normal pixels into faulty pixels and thereby increasing the image’s quality substantially. Experimental results demonstrated than such a scheme is highly effective. Under low bad-pixel density, the Peak-Signal to Noise Ratio (PSNR) can be boosted from the start-of-the-art 41dB to more than 48dB for benchmark image Lena.

    Contents Abstract…………………………………………………….…………..04 Contents………………………………………………………………...05 List of Figures…………………………………………………….……07 List of Tables…………………………………………………………...11 Chapter 1 Introduction………………………………………………..12 1.1 Motivation……………………………………………………13 1.2 Thesis Organization…………………………………….…....14 Chapter 2 Preliminary……………………………………………..….15 2.1 Repair Image without Detector……………………...……….16 2.1.1 Median Filter (MED)……………………………………16 2.1.2 Center Weighted Median Filter (CWM)………….……..16 2.2 Repair Image with Detector……………………………….....17 2.2.1 Signal-Dependent Rank Order Mean (SD-ROM)…….....17 2.2.2 Multiple Thresholds (MTS)……………………………...18 2.2.3 Selective and Adaptive Image Filter (SAIF)…………….19 Chapter 3 Proximity Triangle for Color Image Filter…………….…21 3.1 The Abstract of Proximity Triangle……………………….….21 3.2 Proximity Vector……………………………………………...22 3.3 Proximity Triangle…………………………………………....23 3.4 Computation and Reduction………………………………….24 3.5 Analysis……………………………………………………....24 3.6 False Candidate Problem……………………………………..26 Chapter 4 Proximity Pyramid for Improving Proximity Triangle…28 4.1 The Abstract of Monochromatic Operation…………………..28 4.2 Computation of Monochromatic Operation………………….29 4.3 Variable Threshold……………………………………….…...30 4.4 Repaired Scheme……………………………………….…….31 Chapter 5 Experimental Result………………………………….……32 5.1 Peak-Signal to Noise Ratio (PSNR)…………………….……32 5.2 Fault Model: Salt-Pepper and Its Result………………….…..33 5.3 Fault Model: Variation and Its Result…………………….…..40 5.4 Fault Model: Random Value and Its Result……………….….41 5.5 Run Time…………………………….……………………….47 Chapter 6 Conclusion…………………………………………….……48 Bibliography…………………………………………………….……...49 List of Figures Fig. 1.1: One color pixel = Red cell + Green cell + Blue cell….……….13 Fig. 1.2 (a): The original Lena image………………………….….…….13 Fig. 1.2 (b): 1% bad-cell density in Lena……………………………….13 Fig. 2.1: Classify the methods of image repaired……………………….15 Fig. 2.2: The architecture of method that repair image without detector.16 Fig. 2.3: Addressing 9 image pixels in 3×3 window……………………16 Fig. 2.4: The architecture of method that repair image with detector…..17 Fig. 2.5: An example of operation with SD-ROM…………………...…18 Fig. 2.6: An example of operation with MTS…………………………..19 Fig. 2.7: The four quadrants that represent environment level…………20 Fig. 2.8: The variable threshold of SAIF…………………………….…20 Fig. 3.1: An example of our abstract in color operation……………..…21 Fig. 3.2 (a): An example of the address for color Red…………….……22 Fig. 3.2 (b): An example of proximity vector for Red………………….22 Fig. 3.3 (a): A normal proximity triangle…………………………….....23 Fig. 3.3 (b): An example of bad-cell detection criteria of proximity triangle……………………………………………………..23 Fig. 3.4 (a): An example of characteristic of normal proximity triangle……………………………………………………..25 Fig. 3.4 (b): An example of characteristic of faulty proximity triangle…25 Fig. 3.4 (c): Combine Fig. 3.4(a) and Fig. 3.4(b)…………………….…25 Fig. 3.4 (d): A real case of characteristic of proximity triangle…………25 Fig. 3.5: An example of false candidate problem……………………….27 Fig. 4.1: The idea of our monochromatic operation…………………….29 Fig. 4.2: The variable threshold in our monochromatic operation……...31 Fig. 4.3: An example of how to repair image………………………..….31 Fig. 5.1: Photo-electrical characteristic of image sensor in Salt-Pepper model………………………………………………...…….34 Fig. 5.2 (a): Repaired image with our method in 0% bad-cell density….37 Fig. 5.2 (b): Repaired image with our method in 0.1% bad-cell density...................................................................................37 Fig. 5.2 (c): Repaired image with our method in 1% bad-cell density….37 Fig. 5.2 (d): Repaired image with our method in 5% bad-cell density….37 Fig. 5.3 (a): Lena color image injecting 1% bad-cell density……….…..38 Fig. 5.3 (b): Repair Fig 5.3(a) by MED………………………………....38 Fig. 5.3 (c): Repair Fig 5.3(a) by CWM…………………………..…….38 Fig. 5.3 (d): Repair Fig 5.3(a) by SD-ROM……………………….……38 Fig. 5.3 (e): Repair Fig 5.3(a) by SAIF……………………………...….38 Fig. 5.3 (f): Repair Fig 5.3(a) by MTS………………………….………38 Fig. 5.3 (g): Repair Fig 5.3(a) by our method……………………….….38 Fig. 5.4 (a): Repair Lena image injecting 1% bad-cell density……...….39 Fig. 5.4 (b): Repair Boat image injecting 1% bad-cell density………....39 Fig. 5.4 (c): Repair Goldhill image injecting 1% bad-cell density….…..39 Fig. 5.4 (d): Repair Baboon image injecting 1% bad-cell density……...39 Fig. 5.4 (e): Repair Airplane image injecting 1% bad-cell density……..39 Fig. 5.4 (f): Repair Peppers image injecting 1% bad-cell density………39 Fig. 5.5: Photo-electrical characteristic of image sensor in Variation model………………………………………………………40 Fig. 5.6: The result of MTS and ours in Variation model……………….41 Fig. 5.7: Photo-electrical characteristic of image sensor in Random Value model………………………………………………………41 Fig. 5.8 (a): Repaired image with our method in 0% bad-cell density.…44 Fig. 5.8 (b): Repaired image with our method in 0.1% bad-cell density……………………………………………………..44 Fig. 5.8 (c): Repaired image with our method in 1% bad-cell density….44 Fig. 5.8 (d): Repaired image with our method in 5% bad-cell density.…44 Fig. 5.9 (a): Lena color image injecting 1% bad-cell density……….…..45 Fig. 5.9 (b): Repair Fig 5.9(a) with MED……………………………….45 Fig. 5.9 (c): Repair Fig 5.9(a) with CWM…………………...………….45 Fig. 5.9 (d): Repair Fig 5.9(a) with SD-ROM……………………..……45 Fig. 5.9 (e): Repair Fig 5.9(a) with SAIF……………………………….45 Fig. 5.9 (f): Repair Fig 5.9(a) with MTS………………………………..45 Fig. 5.9 (g): Repair Fig 5.9(a) with our method……………………...…45 Fig. 5.10 (a): Repair Lena image injecting 1% bad-cell density………..46 Fig. 5.10 (b): Repair Boat image injecting 1% bad-cell density………..46 Fig. 5.10 (c): Repair Goldhill image injecting 1% bad-cell density…….46 Fig. 5.10 (d): Repair Baboon image injecting 1% bad-cell density…….46 Fig. 5.10 (e): Repair Airplane image injecting 1% bad-cell density……46 Fig. 5.10 (f): Repair Peppers image injecting 1% bad-cell density…….46 List of Tables Table 1: PSNR is related to the difference between repaired image and original image……………………………………….……..33 Table 2: The results of different methods in different bad-cell density in Salt-Peppers model………………………………………...35 Table 3: The total numbers of normal cells over-detected in different methods in Salt-Peppers model……………………………36 Table 4: The results of MTS and ours with different test images in Salt-Pepper model…………………………………………36 Table 5: The results of different methods in different bad-cell density in Random Value model………………………………...…….42 Table 6: The total numbers of normal cells over-detected in different methods in Random Value model………………………….43 Table 7: The results of SD-ROM and ours with different test images in Random Value model………………………………...…….43 Table 8: The run time is shown in different methods…………...………47

    Bibliography

    [1] A. Bovik, “Handbook of Image and Video Processing,” Boston, MA :Academic Press, pp.11-12, 2005.
    [2] A. Ceremonies, F. Maloberti, G. Torelli, and C. Vacchi, “An 8-Bit Two-Step Flash A/D Converter for Video Applications,” IEEE CICC. pp. 6.3/1-6.3/4, May 1989.

    [3] A. Yukawa, ”A CMOS 8-Bit High-Speed A/D Converter IC,” IEEE Journal of Solid-State Circuits, vol. 20, no. 3, pp. 775-779, June 1985.

    [4] E. Abreu and S. K. Mitra, “A Signal-Dependent Rank Order Mean (SD-ROM) Filter. A New Approach for Removal of Impulses from Highly Corrupted Images,” in Proc. IEEE ICASSP’95, Detroit, MI, pp. 2371-2374, May 1995.

    [5] E. R. Fossum, “CMOS Image Sensors: Electronic Camera on A Chip,” Int. Electron Device Meeting 1995, pp. 17-25, Dec. 1995.

    [6] F. Russo, “An Image Enhancement Technique Combining Sharpening and Noise Reduction,” IEEE Transaction on Instrumentation and Measurement, vol.51, no. 8, pp. 1148-1154, Aug. 2002.

    [7] F. Russo, “A Method for Estimation and Filtering of Gaussian Noise in Images,” IEEE IEEE Transaction on Instrumentation and Measurement, vol. 52, no. 8, pp. 1148-1154, Aug. 2003.

    [8] G. K. Robert, “Cubic Convolution Interpolation for Digital Image Processing,” IEEE Transactions on Acoustics, Speech and Signal Processing, vol. 29, no. 6, pp. 1153-1160, Dec. 1981.

    [9] H. Hsieh and C. A. Harry, “Cubic Splines for Image Interpolation and Digital Filtering,” IEEE Transaction on Acoustics, Speech, and Signal Processing, vol. 26, no. 6, pp. 508-517, Dec. 1978.

    [10] H. M. Lin and N. W. Alan, “Median Filters with Adaptive Length,” IEEE Transaction on Circuits and Systems, Vol. 6, pp. 675-690, June 1988.

    [11] H. Wei and T.D. Binnie, “High-Resolution Image Reconstruction from Multiple Low-Resolution Images,” International Conference on Image Processing and its Applications, 1996, vol. 2, pp. 596-600, July 1999

    [12] K. J. Kerpez, “Minimum Mean Squared Error Impulse Noise Estimation and Cancellation,” IEEE Transactions on Signal Processing, vol. 43, pp.1651-1662, July 1995.

    [13] K. R. Castleman, “Digital Image Processing,” Englewood Cliffs, N.J.: Prentice Hall, pp. 547-551, 1996.

    [14] L. Bar, A. Brook, N. Sochen, and N. Kiryati, “Deblurring of Color Images Corrupted by Impulsive Noise,” IEEE Transactions on Image Processing, vol. 16, pp. 1101-1111, April 2007.

    [15] S.J. Ko and Y.H. Lee, ”Center Weighted Median Filter and Their Applications to Image Enhancement,” IEEE Transaction on Circuits and Systems, vol. 38, pp. 984-993, Sept. 1991.

    [16] S. S. H. Naqvi, N. C. Gallagher, and E. J. Coyle, “An Application of Median Filter to Digital Television,” in Proc. IEEE ICASSP’86, vol. 11, pp. 2451-2454, April 1986.

    [17] T. C. Lin and P. T. Yu, ”Salt-Pepper Impulse Noise Detection and Removal Using Multiple Thresholds for Image Restoration,” Journal of Information Science and Engineering, vol. 22, pp. 189-198, Jan. 2006.

    [18] W. C. Kao and Y. J. Chen, “Multistage Bilateral Noise Filtering and Edge Detection for Color Image Enhancement,” IEEE Transactions on Consumer Electronics, vol. 51, pp. 1346-1351, Nov. 2005.

    [19] X. Han, J. Li, Y. Li, and X. Xu, ”A Selective and Adaptive Image Filtering Approach Based on Impulse Noise Detection,” in Proc. WCICA 2004, vol. 5, pp. 4156-4159, June 2004.

    [20] Y. H. Lee, “Generalized Median Filtering and Related Nonlinear Filtering Technique,” IEEE Transactions on Acoustics, Speech and Signal Processing, vol. 33, no. 3, pp. 672-683, June 1985.

    無法下載圖示 全文公開日期 本全文未授權公開 (校內網路)
    全文公開日期 本全文未授權公開 (校外網路)

    QR CODE