研究生: |
謝德偉 De-Wei Hsieh |
---|---|
論文名稱: |
利用CNN 模型對老舊竣工圖進行有效的二值化 Novel and Effective CNN-based Binarization for Historically Degraded Asbuilt Drawing Maps |
指導教授: |
鍾國亮
Kuo-Liang Chung |
口試委員: |
花凱龍
Kai-Lung Hua 金台齡 Tai-lin Chin 陳建中 Jiann-Jone Chen |
學位類別: |
碩士 Master |
系所名稱: |
電資學院 - 資訊工程系 Department of Computer Science and Information Engineering |
論文出版年: | 2021 |
畢業學年度: | 109 |
語文別: | 中文 |
論文頁數: | 26 |
中文關鍵詞: | 深度學習 、影像處理 、二值化 、竣工圖 |
外文關鍵詞: | deep learning, image process, binarization, as-built map |
相關次數: | 點閱:180 下載:8 |
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 |
對於有歷史且老舊的竣工圖進行二值化是一項新的具有挑戰性的工作,尤其是在消除三個因環境或時間所造成的瑕疵(即雜點,泛黃區域和摺痕)的同時,還要很好地保留住清楚的前景。在本文中,我們提出了一種新的深度學習二值化方法,以生成高品質的二值化竣工圖。基於該方法所測試出來的二值化圖,我們的實驗數據可看出,就精確度、PSNR(峰值信號噪聲比)和其視覺效果而言,我們的方法優於現有的九種二值化方法,其中也包含了兩種基於深度學習所提出的二值化演算法,且我們所提出的深度學習模型也是三種深度學習模型裡最快速且參數量最少的。
Binarizing historically degraded asbuilt drawing maps(HDAD) is a new challenging job, especially in terms of removing the three artifacts, namely noise, the yellowing areas, and the folded lines, while preserving the foreground components well. In this paper, we first propose a convolutional neural networkbased (CNNbased) binarization method to produce highquality binarized HDAD maps. Based on the testing HDAD maps, the thorough experimental data demonstrated that in terms of the accuracy, PSNR (peaksignal tonoiseratio), and the perceptual effect of the binarized HDAD maps, our method substantially outperforms the nine existing binarization methods. In addition, with similar accuracy, the experimental results demonstrated the significant executiontime reduction merit of our method relative to the retrained version of the stateoftheart CNNbased binarization methods.
[1] A. S. Abutaleb, “Automatic thresholding of graylevel
pictures using twodimensional
entropy,”Computer Vision, Graphics, and Image Processing, vol. 47,
no. 1, pp. 22–32, 1989.
[2] V. Badrinarayanan, A. Kendall, and R. Cipolla, “Segnet: a deep convolutional
encoderdecoder
architecture for image segmentation,” IEEE Transactions on Pattern
Analysis and Machine, vol. 39, no. 12, pp. 24812495,
2017.
[3] J. Bernsen, “Dynamic thresholding of graylevel
images,”Proceedings of International
Conference on Pattern Recognition (ICPR), Paris, France, 1986, pp. 1251–
1255.
[4] J. CalvoZaragoza,
G. Vigliensoni, I. Fujinaga, “Pixelwise
binarization of musical
documents with convolutional neural networks,” International Conference on Machine
Vision Applications (ICMVA), Nagoya, Japan, 2017, pp. 362365.
[5] J. CalvoZaragoza
and A. J. Gallego, “A selectional autoencoder
approach for document
image binarization,” Pattern Recognition, vol. 86, no. 2, pp. 3747,
2019.
[6] Q. Chen, Q. S. Sun, P. A. Heng, and D. S. Xia, “A doublethreshold
image binarization
method based on edge detector,” Pattern Recognition, vol. 41, no. 4, pp.
12541267,
2008.
[7] Y. H. Chiu, K. L. Chung, W. N. Yang, Y. H. Huang, and C. H. Liao, “Parameterfree
based twostage
method for binarizing degraded document images,” Pattern
Recognition, vol. 45, no. 12, pp. 42504262,
2012.
[8] K. L. Chung, D. W. Hsieh, and C. H. Liao, “Effective binarization for historically
degraded asbuilt
drawing maps using convolutional neural networks,” International
Workshop on Advanced Image Technology (IWAIT), Yogyakarta, Indonesia, 2020,
vol. 11515, pp. 3035.
[9] V. Dumoulin and F. Visin, “A guide to convolution arithmetic for deep learning,”
arXiv: 1603.07285, 2016.
[10] DIBCO dataset https://vc.ee.duth.gr/dibco2019/
[11] ftp://140.118.175.164/HDAD_maps.
[12] B. Gatos, I. Pratikakis, and S. J. Perantonis, “Adaptive degraded document image
binarization,” Pattern Recognition, vol. 39, no. 3, pp. 317327,
2006.
[13] R. C. Gonzalez and R. E. Woods, “Digital Image Processing,” 4th Edition, Pearson,
NJ, USA, 2017.
[14] A. A. Gooch, S. C. Olsen, J. Tumblin, and B. Gooch, “Color2Gray: Saliencepreserving
color removal,”ACM Transactions on Graphics, vol. 24, no. 3, pp. 634–
639, 2005.
[15] M. Grundland and N. A. Dodgson, “Decolorize: Fast, contrast enhancing, color to
grayscale conversion,”Pattern Recognition, vol. 40, no. 11, pp. 2891–2896, 2007.
[16] R. Hedjam, H. Z. Nafchi, M. Kalacska, and M. Cheriet, “Inuence of colortogray
conversion
on the performance of document image binarization: toward a noveloptimization
problem,” IEEE Transactions. Image Processing, vol. 24, no. 11, pp.
3637–3651, 2015.
[17] N. R. Howe, “Document binarization with automatic parameter tuning,” International
Journal on Document Analysis and Recognition, vol. 16, no. 3, pp. 247258,
2012.
[18] K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,”
The IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
Las Vegas, NV, USA, 2016, pp. 770778.
[19] F. Jia, C. Shi, K. He, C. Wang, and B. Xia, “Degraded document image binarization
using structural symmetry of strokes,” Pattern Recognition, vol. 74, no. 2, pp. 225240,
2018.
[20] E. Kavallieratou, “A binarization algorithm specialized on document images and
photos,” Proceedings of the 8th International Conference on Document Analysis and
Recognition (ICDAR), Washington, DC, USA, 2005, pp. 463467.
[21] J. Kittler and J. Illingworth, “On threshold selection using clustering criteria,” IEEE
Transactions on Systems, Man, and Cybernetics, vol. 15, no. 5, pp. 652655,
1985.
[22] S. J. Ko and Y. H. Lee, “Center weighted median filters and their applications to
image enhancement,” IEEE Transactions on Circuits and Systems, vol. 38, no. 9, pp.
984–993, 1991.
[23] A. Krizhevsky, I. Sutskever, and G. Hinton, “Imagenet classification with deep convolutional
neural networks,” Advances in neural information processing systems, pp.
10971105,
2012.
[24] J. N. Kapur, P. K. Sahoo, and A. K. C. Wong, “A new method for gray level picture
thresholding using the entropy of histogram,” Computer Vision, Graphics, and Image
Processing, vol. 29, no. 3, pp. 273285,
1985.
[25] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic
segmentation,” IEEE Conference on Computer Vision and Pattern Recognition
(CVPR), Boston, MA, USA, 2015, pp. 34313440.
[26] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no.7553,
pp. 436444,
2015.
[27] S. Lu, B. Su, and C. L. Tan, “Document image binarization using background estimation
and stroke edges,”International journal on document analysis, vol. 13, no.
4, pp. 303–314, 2010.
[28] R. F. Moghaddam and M. Cheriet, “A multiscale
framework for adaptive binarization
of degraded document images,” Pattern Recognition, vol. 43, no. 6, pp. 21862198,
2010.
[29] W. Niblack, “An introduction to digital image processing,” PrenticeHall,
Englewood
Cliffs, NJ, pp. 115116,
1986.
[30] N. Otsu, “A threshold selection method from gray level histograms,” IEEE Transactions
on Systems, Man, and Cybernetics, vol. 9, no. 1, pp. 6266,
1979.
[31] J. PastorPellicer,
S. E. Boquera, F. ZamoraMartnez,
M. Z. Afzal, and M. J. C.
Bleda, “Insights on the use of convolutional neural networks for document image
binarization,” International WorkConference
on Artical Neural Networks (IWANN),
Palma de Mallorca, Spain, 2015, no. 2, pp. 115126.
[32] O. Ronneberger, P. Fischer, and T. Brox, “Unet:
Convolutional networks for
biomedical image segmentation,” Medical Image Computing and Computer Assisted
Intervention (MICCAI), Munich, Germany, 2015, pp. 234–241.
[33] J. Sauvola and M. Pietikäinen, “Adaptive document image binarization,” Pattern
Recognition, vol. 33, no. 2, pp. 225236,
2000.
[34] K. Smith, P. E.
Landes, J. Thollot, and K. Myszkowski, “Apparent greyscale: A
simple and fast conversion to perceptually accurate images and video,”Computer
Graphics Forum, vol. 27, no. 2, pp. 193–200, 2008.
[35] B. Su, S. Lu, and C. L. Tan, “Binarization of historical handwritten document images
using local maximum and minimum filter,”International Workshop on Document
Analysis Systems (DAS), New York, NY, USA, 2010, pp. 159–166.
[36] B. Su, S. Lu, and C. L. Tan, “Robust document image binarization technique for
degraded document images,” IEEE Transactions on image processing, vol. 2, no. 4,
pp. 14081417,
2013.
[37] C. Tensmeyer and T. Martinez, “Document image binarization with fully convolutional
neural networks,” Proceedings of the 14th International Conference on Document
Analysis and Recognition (ICDAR), Kyoto, Japan, 2017, no. 1, pp. 99104.
[38] Q. N. Vo, S. H. Kim, H. J. Yang, and G. Lee, “Binarization of degraded document
images based on hierarchical deep supervised network,” Pattern Recognition, vol.
74, pp. 568586,
2018.
[39] J. H. Xue and D. M. Titterington, “tTest,
Ftests
and Otsu’s methods for image
thresholding,” IEEE Transactions on image processing, vol. 20, pp. 23922396,
2011.
[40] M. Zhao, Y. Yang, and H. Yan, “An adaptive thresholding method for binarization of
blueprint images,” Pattern Recognition Letters, vol. 21, no. 10, pp. 927943,
2000.
[41] J. Zhao, C. Shi, and F. Jia, “Document image binarization with cascaded generators
of conditional generative adversarial networks,” Pattern Recognition, vol. 96, no.
12, 2019.