视觉乐谱图像动态多阈值二值化方法

doi:10.11896/j.issn.1002-137X.2016.01.067

摘要/Abstract

摘要： 针对视觉乐谱图像在不规则区域内光照不均匀的二值化问题,提出了一种对前景内容动态分块二值化的方法,以降低背景噪声对二值化效果的影响,提高分块多阈值方法的有效性。该方法以像素灰度差分为基础,以滑动窗口为处理对象,根据窗口内差分变化率确定前景区域位置；再根据前景区域背景像素灰度分布情况将窗口分类,从而划分出不规则光照区域；最后在每个区域内分别采用Otsu阈值算法进行二值化。在实验中详细讨论了窗口大小和差分变化率阈值设置对前景区域提取正确率的影响,结果表明所提出的方法对乐谱图像前景内容的查全率高,能划分出不规则形状的范围,可以有效处理光照不均匀情况,并且可以用于一般文档的二值化。

关键词: 二值化,乐谱图像,差分,前景区域检测,动态阈值

Abstract: A dynamic binarization method was proposed for solving the uneven illumination problems in irregular regions of visual music images,to reduce the background influence and improve the effectiveness of multi-threshold me-thod.This method is based on pixel grayscale difference,and slide window is the segment unit.The foreground is detected by the difference variance ratio in each window,and the irregular illumination regions are decided by window background gray estimation.The Otsu’s binarization threshold is employed in each region in this algorithm.Foreground detection accuracy is decided by window size and variance ratio threshold.Those two parameters were discussed in detail in experiment.The experimental results show that this method can divide foreground into irregular illumination regions with a high foreground recall ratio.And it is effective to solve the uneven illumination binarization problem of music images.This method can also be applied to the binarization of general document images.

Key words: Binarization,Music image,Difference,Foreground detection,Dynamic threshold

杨芳,滕桂法,田学东. 视觉乐谱图像动态多阈值二值化方法[J]. 计算机科学, 2016, 43(1): 310-314. https://doi.org/10.11896/j.issn.1002-137X.2016.01.067

YANG Fang, TENG Gui-fa and TIAN Xue-dong. Dynamic Multi-threshold Binarization Method for Visual Music Images[J]. Computer Science, 2016, 43(1): 310-314. https://doi.org/10.11896/j.issn.1002-137X.2016.01.067

参考文献

[1] Jagroop K D,Mahanan R.A review of degraded document image binarization techniques[J].International Journal of Advanced Research in Computer and Communication Engineering,2014,3(5):6582-6586
[2] Wang J F,Huang Z C,Ahmed T A M.New binarization method called BM aim to optimize detail of image[J].Journal of Wuhan University of Technology,2014,6(8):127-132(in Chinese)王骏飞,黄樟灿,Ahmed T A M.一种针对图像细节优化的BM二值化方法[J].武汉理工大学学报,2014,36(8):127-132
[3] Howe N R.Document binarization with automatic parametertuning[J].International Journal on Document Analysis and Re-cognition,2013,16(3):247-258
[4] Singh T R,Roy S,Singh O I,et al.A new local adaptive thre-sholding technique in binarization[J].International Journal of Computer Science Issues,2011,8(6):271-277
[5] Nafchi H Z,Moghaddam R F,Cheriet M.Historical document binarization based on phase information of images[C]∥Compu-ter Vision-ACCV 2012 Workshops.Springer Berlin Heidelberg,2013:1-12
[6] Hu L N.Study on low quality document image binarization [D].Nanjing:Nanjing University of Technology and Engineering,2012(in Chinese)胡丽娜.低质量文档图像的二值化研究[D].南京:南京理工大学,2012
[7] Zhang W,Liu Z G.Binarization method to scanning image of thicker book [J].Application Research of Computers,2011,8(5):1998-2000(in Chinese)张伟,刘志刚.一种针对较厚书籍扫描图像的二值化方法[J].计算机应用研究,2011,28(5):1998-2000
[8] Lan H.Interactive medical image segmentation algorithm optimized by multi-thresholds [J].Computer Science,2013,40(9):296-299(in Chinese)兰红.多阈值优化的交互式医学图像分割方法[J].计算机科学,2013,40(9):296-299
[9] Valizadeh M,Armanfard N,Komeili M,et al.A novel hybrid algorithm for binarization of badly illuminated document images[C]∥14th International Computer Conference on CSI (CSICC 2009).IEEE,2009:121-126
[10] He Z M.Binarization of document image captured under uneven illumination [J].Journal of Shanghai University of Engineering Science,2011,5(2):163-166(in Chinese) 贺志明.不均匀光照文本图像的二值化[J].上海工程技术大学学报,2011,5(2):163-166
[11] Ma C Y,Sun J X,Cheng S,et al.Binarization algorithm of license plate image based on multistage light processing[J].Journal of Computer Applications,2013,3(S2):200-202(in Chinese) 马超玉,孙俊喜,程帅,等.融合多级光照处理的车牌图像二值化算法[J].计算机应用,2013,3(S2):200-202
[12] Burgoyne J A,Pugin L,Eustace G,et al.A comparative survey of image binarisation algorithms for optical recognition on degraded musical sources[C]∥International Society for Music Information Retrieval Conference (ISMIR).2007:509-512
[13] Sauvola J,Seppanen T,Haapakoski S,et al.Adaptive document binarization[C]∥Proceedings of the Fourth International Conference on Document Analysis and Recognition(ICDAR).IEEE.1997:147-152
[14] Vigliensoni G,Burlet G,Fujinga I.Optical measure recognition in common music notation[C]∥Proceedings of International Symposium/Conference on Music Information Retrieval (ISMIR).2013:125-130
[15] Pinto T,Rebelo A,Giraldi G,et al.Music score binarizationbased on domain knowledge[M]∥Pattern Recognition and Ima-ge Analysis.Springer Berlin Heidelberg,2011:700-708
[16] Guo J,Liu X Y,Wu B,et al.Bniarisation method for images acquired under non-uniform illumination [J].Computer Applications and Software,2014,31(3):183-188(in Chinese) 郭佳,刘晓玉,吴冰,等.一种光照不均匀图像的二值化方法[J].计算机应用与软件,2014,31(3):183-188
[17] Su B,Lu S,Tan C L.Robust document image binarization technique for degraded document images [J].IEEE Transactions on Image Processing,2013,22(4):1408-1417
[18] Xu H Y,Ma L L,Wu J.Document image binarization based on background estimation and edge detection [J].Computer Applications and Software,2014,31(8):196-200(in Chinese) 许海洋,马龙龙,吴健.基于背景估计和边缘检测的文档图像二值化[J].计算机应用与软件,2014,31(8):196-200
[19] Zhang J R,Wang L J.Improved binarization algorithm of modified QR code image [J].Modern Electronics Technique,2014,7(7):56-58(in Chinese) 张继荣,王利军.一种改进的QR码图像二值化算法[J].现代电子技术,2014,7(7):56-58

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed