在這一篇論文中，我們提出利用比較原始與重建半色調影像之半色調轉連續調影像的方法。當兩張連續調的影像很相似的時候，它們的半色調影像會相同。
一張經過反轉半色調演算法還原的重建連續調影像應該和原始連續調影像很相似，此外，當這兩張連續調影像的半色調影像相同時，這兩張連續調影像會更相似。因此，在本篇論文中，我們利用原始半色調和重建半色調影像的差異來對重建連續調影像作調整。本篇論文中所提及的半色調影像都是由誤差擴散法所產生的，而重建連續調影像則是由查表反轉半色調法所產生。
本篇論文我們提出兩個調整重建連續調影像的方法。第一個方法利用目前所在位置像素和它後面一個像素的半色調來對重建連續調影像做修正，改根據我們的觀察，兩個像素的半色調影像和連續調影像上有很類似的趨勢，我們會一直調整相對應的重建連續調影像的值直到重建半色調影像和原半色調影像相同為止。第二個方法則是利用事先計算好的查表來產生補償值，為了補償結果的效能，我們利用K-means群集演算法將影像做分類，每類表格的對應方式是將兩張半色調影像的差值對應到相對應的原始連續調影像和重建連續調影像的差值。實驗結果顯示我們所提出來的方法在效能上比其他方法好。

In this paper, an inverse half-toning method based on the comparison between an original halftone and its re-generated halftone is proposed. Two halftone images coded by a half-toning algorithm are the same while their continuous-tone (contone) images are far more similar. A reconstructed contone image generated by an inverse half-toning algorithm should be similar to its original one. Moreover, the two contone images would be more similar if their halftones are the same. In this paper, therefore, the pixel value of a reconstructed contone image is adjusted according to the difference between its original halftone and its new one. All halftones in the paper are generated by error-diffusion method and the reconstructed contones are produced by lookup table method. Two kinds of adjusting methods are proposed in the paper. The first is to modify two pixel’s contone values by using the halftones of the pixel and its latter one. According to our observations, the tendency between the halftones and contones of the two pixels is similar, and their contone values are adjusted at the same time until their re-generated halftones are the same as their original ones. The second is to generate compensative values by using pre-computing lookup tables. In order to improve the performance, an image is classified into several categories by K-means algorithm. The table of every category is mapped from the difference between the two halftones to the corresponding difference between original and reconstructed contone images. Experimental results show that the proposed methods have better performance compared with others.

[1] W. H. F. Talbot, “Improvement in the art of engraying,” British Patent Specification No. 565, October 29, 1852
[2] R.W. Floyd and L. Steinberg, “An adaptive algorithm for spatial grayscale,” in Proc. SID, vol. 17/2, pp. 75-77, 1976
[3] C. N. Judice, J. F. Jarvis, and W. H. Ninke, “Using ordered dither to display continuous tone pictures on AC plasma panel,” Proc, SID, Vol. 15, pp. 161-169, 1947
[4] P. Stucki, “MECCA-a multiple-error correcting computation algorithm for bilevel image hardcopy reproduction,” Research report RZ1060, IBM Research Lab., Zurich, Switzerland, 1981
[5] B. E. Bayer, “An optimum method for two-level rendition of continuous-tone pictures,” Proc. IEEE Int. Conf. Communications, vol. 26, pp. 2611–2615, June 11–13, 1973.
[6] J. R. Sullivan, L. A. Ray, and R. Miller, “Design of minimum visual modulation halftone patterns,” IEEE Trans. Syst., Man, Cybern., vol. 21, pp. 33–38, Jan. 1991.
[7] K.E. Spaulding, R. L. Miller, and J. Schildkraut, “Methods for generating blue-noise dither matrices for digital halftoning”, JEI, Vol. 6, pp. 208-230, Jan. 1997
[8] P.C. Chang, C.S. Yu, and T.H. Lee, "Hybrid LMS-MMSE Inverse Halftoning Technique," IEEE Trans. On Image Processing, vol.10, No.1, p.p.95-103, Jan. 2001
[9] M. Mese and P. P. Vaidyanathan, “Look up table (LUT) method for inverse halftoning,” IEEE Trans. Image Processing, vol. 10, pp. 1566–1578, Apr. 2000.
[10] T. D. Kite, N. Damera-Venkata, B. L. Evans, and A. C. Bovik, “A fast, high-quality inverse halftoning algorithm for error diffused halftones,” IEEE Trans. Image Process., vol. 9, no. 9, pp. 1583–1592, Sep. 2000.
[11] M. Analoui and J. P. Allebach, “New results on reconstruction of continuous-tone from halftone,” in Proc. IEEE Int. Conf. Acoustics, Speech,Signal Processing, vol. 3, pp. 313–316. 1992,
[12] S. Hein and A. Zakhor, “Halftone to continuous-tone conversion of error-diffusion coded images,” in Proc. Int. Conf Acoustics, Speech, and Signal Processing, Minneapolis, MN, pp. V:309-V:312, Apr. 1993
[13] M. Y. Shen and C.-C. J. Kuo, “A robust nonlinear filtering approach to inverse halftoning,” J. Vis. Commun. Image Represen., vol. 12, pp. 84–95, Mar. 2001.
[14] T. D. Kite, N. Damera-Venkata, B. L. Evans, and A. C. Bovik, “A fast, high-quality inverse halftoning algorithm for error diffused halftones,” IEEE Trans. Image Processing, vol. 9, pp. 1583–1592, Sept. 2000.
[15] N. Damera-Venkata, T. D. Kite, M. Venkataraman, and B. L. Evans, “Fast blind inverse halftoning,” in Proc. IEEE Conf. Image Processing, pp. 64–68.　Oct. 1998,
[16] Y. Kim, G. Arce, and N. Grabowski, “Inverse halftoning using binary permutation filters,” IEEE Trans. Image Processing, vol. 4, pp. 1296–1311, Sept. 1995.
[17] R. S. Kern, T. G. Stockham Jr., and D. C. Strong, “Descreening via linear filtering and iterative techniques,” in Proc. SPIE Conf. Human Vision,Visual Processing, Digital Display IV, vol. 1913, San Jose, CA, pp. 299–309. Feb. 1993,
[18] J. Luo, R. de Queiroz, and Z. Fan, “A robust technique for image descreening based on the wavelet transform,” IEEE Trans. Signal Processing, vol. 46, pp. 1179–1194, Apr. 1998.
[19] Z. Xiong, M. Orchard, and K. Ramchandran, “Inverse halftoning using wavelets,” in IEEE Trans. Image Processing, pp. 1479–1483. Oct. 1999,
[20] M. Mese and P. P. Vaidyanathan, “Template selection for LUT inverse halftoning and application to color halftones,” Proc. IEEE ICASSP’00, vol. 4, pp. 2290–2293, June 2000
[21] K.L.Chung and S.T.Wu, Inverse halftoning algorithm using edge-based lookup table approach, IEEE Trans. on Image Processing, pp. 1583-1589 14(10), 2005,
[22] Y. Mita, S. Sugiura, and Y. Shimomura, “High quality multi-level image restoration from bi-level image,” in Proc. Int. Congr. Advances Non-Impact Printing Technologies Black-and-White Color, Orlando, FL, pp. 791–802. Oct. 21–26, 1990,
[23] H. Y. Kim and R. L. Queiroz, “Inverse Halftoning by Decision Tree Learning,” in Proc. IEEE Int. Conf. on Image Processing, (Barcelona, Spain), vol. 2, pp. 913-916, 2003
[24] J. Z. C. Lai and J. Y. Yen, “Inverse error-diffusion using classified vector quantization,” IEEE Trans. Image Processing, vol. 7, pp. 1753–1758, Dec. 1998.
[25] S. M. Schweizer and R. L. Stevenson, “A Bayesian approach to inverse halftoning,” in Proc. SPIE Conf. Human Vision, Visual Processing, Digital Display IV, vol. 1913, San Jose, CA, Feb. 1993, pp. 282–292.
[26] R. L. Stevenson, “Inverse halftoning via MAP estimation,” IEEE Trans.Image Processing, vol. 6, pp. 574–583, Apr. 1997.
[27] [Online]. Available: http://www.systems.caltech.edu/mese/halftone
[28] J. B. MacQueen, "Some Methods for classification and Analysis of Multivariate Observations, Proceedings of 5-th Berkeley Symposium on Mathematical Statistics and Probability", Berkeley, University of California Press, 1:281-297 1967
[29] W. B. Huang, W. C. Chang, Y. W. Lu, W.Y. Su, Y. H. Kuo, “Halftone/Contone Conversion Using Neural Networks” International Conference on Image Processing, vol. 5, pp 3547- 3550, 2004
[30] T. N. Pappas and D. L. Neuhoff, “Least-squares model-based halftoning,” in Proc. 1992 SPIE/IS&T Symp. Electronic Imaging Science Technology, vol. 1666, San Jose, CA, pp. 165–176. Feb. 1992,
[31] M. Analoui and J. P. Allebach, “Model-based halftoning using direct binary search,” in Proc. IS&T/SPIE Int. Symp. on Electronic Imaging Science and Technology, vol. 1666, San Jose, CA, pp. 96–108. Feb. 9–14, 1992,