基于Shearlet变换的扫描文档图像倾斜检测
|
摘要
针对含有文本、图像、表格和公式等复杂扫描电子文档的倾斜校正问题,提出了一种基于Shearlet(剪切波)变换与多尺度分析的复杂文档图像倾斜检测方法,利用剪切波变换的局部性和方向性,可以得到各个方向的能量值,能量最高的两个分量分别对应于横画与竖画的方向,根据这两个方向可以较准确地检测出扫描文档图像中文字行的方向并进一步确定扫描文档图像的倾斜角度。实验结果表明,该方法可以避免扫描文档中图像、表格等的干扰,具有较好的抗噪声能力,对于具有复杂内容的文档图像有较高的检测准确率。
For the skew correction of complex scanned electronic documents with text,images,tables and formulas,a new skew detection method is proposed based on Shearlet transform and multi-scale analysis.Based on the localization and directionality of Shearlet transform,energy in all directions can be obtained.The two components with the highest energy can correspond to the horizontal and vertical directions,respectively.According to these two directions,the direction of the text line in the scanned document can be accurately detected and the skew angle of document image can be further determined.The experimental results show that the proposed method can avoid the interference of images,tables,and so on in scanned documents,and has a good anti-noise ability.The proposed method is of higher detection accuracy for document images with complex content.
For the skew correction of complex scanned electronic documents with text,images,tables and formulas,a new skew detection method is proposed based on Shearlet transform and multi-scale analysis.Based on the localization and directionality of Shearlet transform,energy in all directions can be obtained.The two components with the highest energy can correspond to the horizontal and vertical directions,respectively.According to these two directions,the direction of the text line in the scanned document can be accurately detected and the skew angle of document image can be further determined.The experimental results show that the proposed method can avoid the interference of images,tables,and so on in scanned documents,and has a good anti-noise ability.The proposed method is of higher detection accuracy for document images with complex content.
引文
[1]Singh B,Maini R.Skew detection and correction of Gurmukhi words from natural scene images[J].International Journal of Signal Processing,Image Processing and Pattern Recognition,2016,9(9):139-146.
[2]LiuQ J,Zhao Z M,Zhang W J.Study on skew correction method for gene-chip scanning image based on pixel gray[J].Optical Technique,2017,43(2):108-113.刘全金,赵志敏,张文杰.一种基因芯片光学扫描图像倾斜校正方法研究[J].光学技术,2017,43(2):108-113.
[3]Liao Z,Qiu Q,Zhang Y D.Numerical simulation of segmented telescope[J].Acta Optica Sinica,2014,34(7):0722002.廖周,邱琪,张雨东.分块拼接望远镜的数值仿真[J].光学学报,2014,34(7):0722002.
[4]Zhang Y F.Research of image denoising and document image skew correction based on Shearlet[D].Kaifeng:Henan University,2016:5-20.张一凡.基于Shearlet变换的图像降噪及文档图像倾斜校正研究[D].开封:河南大学,2016:5-20.
[5]Boukharouba A.A new algorithm for skew correction and baseline detection based on the randomized Hough Transform[J].Journal of King Saud University-Computer and Information Sciences,2017,29(1):29-38.
[6]Ahmad R,Rashid S F,Afzal M Z,et al.A novel skew detection and correction approach for scanned documents[C]//DAS 2016,12th Intl IAPRWorkshop on Document Analysis Systems,Santorini,Greece,2016.
[7]Alaei A,Nagabhushan P,Pal U,et al.An efficient skew estimation technique for scanned documents:An application of piece-wise painting algorithm[J].Journal of Pattern Recognition Research,2016,11(1):1-14.
[8]O'Reilly C,Gluhak A,Imran M A.Distributed anomaly detection using minimum volume elliptical principal component analysis[J].IEEE Transactions on Knowledge and Data Engineering,2016,28(9):2320-2333.
[9]Verma R N,Malik L G.Review of illumination and skew correction techniques for scanned documents[J].Procedia Computer Science,2015,45(1):322-327.
[10]Easley G R,Labate D,Colonna F.Shearlet-based total variation for denoising[J].IEEE Transactions on Image Processing,2009,18(18):260-268.
[11]Jakobsen M,Lemvig J.Reproducing formulas for generalized translation invariant systems on locally compact abelian groups[J].Transactions of the American Mathematical Society,2016,368(12):8447-8480.
[12]Jenifha M A,Babu G M.Nonparametric illumination and skew correction for scanned document images[J].Digital Image Processing,2016,8(5):165-169.
[13]Hu J H.Image denoising based on Shearlet transform directionality[D].Xi′an:Northwest University,2014:14-26.胡江华.基于Shearlet变换方向性的图像消噪[D].西安:西北大学,2014:14-26.
[14]Koo H I,Cho N I.Robust skew estimation using straight lines in document images[J].Journal of Electronic Imaging,2016,25(3):033014.
[15]Wu Y Q,Tao F X.Multispectral and panchromatic image fusion based on improved projected gradient NMF in NSST domain[J].Acta Optica Sinica,2015,35(4):0410005.吴一全,陶飞翔.改进投影梯度NMF的NSST域多光谱与全色图像融合[J].光学学报,2015,35(4):0410005.
[16]Shi M H,Liu W.Anew image denoising algorithm based on shift-invariant Shearlet transform domain[J].Infrared Technology,2016,38(1):33-40.石满红,刘卫.一种新的平移不变Shearlet变换域图像去噪算法[J].红外技术,2016,38(1):33-40.
[17]Guo K,Labate D.Characterization and analysis of edges using the continuous Shearlet transform[J].Siam Journal on Imaging Sciences,2009,2(3):959-986.
[18]Easley G R,Labate D,Lim W Q.Optimally sparse image representations using shearlets[C]//Proceedings of IEEE International Conference on Signals,Systems and Computers,2006:974-978.
[19]Zhang J X.The theory of modern Chinese character pen shape[C].The Selection of the 2nd International Chinese Teaching Symposium.Beijing:Beijing Language Institute Press,1988:162-165.张静贤.现代汉字笔形论[C].第二届国际汉语教学讨论会论文选.北京:北京语言学院出版社,1988:162-165.
[20]Feng Z W.Computer processing of natural language[J].Chinese Information,1997(4):26-27.冯志伟.自然语言的计算机处理[J].中文信息,1997(4):26-27.
[21]Fan L.Research on stroke characteristics and feature of Chinese characters[D].Chongqing:Southwestern University,2013:17-27.范玲.现行汉字笔画分布及特征研究[D].重庆:西南大学,2013:17-27.
[22]Papandreou A,Gatos B,Louloudis G,et al.ICDAR2013document image skew estimation contest[C]//Proceedings of IEEE International Conference on Document Analysis and Recognition,2013:1444-1448.
[23]Chou C H,Chu S Y,Chang F.Estimation of skew angles for scanned documents based on piecewise covering by parallelograms[J].Pattern Recognition,2007,40(2):443-455.
[24]Dey P,Noushath S.e-PCP:A robust skew detection method for scanned document images[J].Pattern Recognition,2010,43(3):937-948.
[2]LiuQ J,Zhao Z M,Zhang W J.Study on skew correction method for gene-chip scanning image based on pixel gray[J].Optical Technique,2017,43(2):108-113.刘全金,赵志敏,张文杰.一种基因芯片光学扫描图像倾斜校正方法研究[J].光学技术,2017,43(2):108-113.
[3]Liao Z,Qiu Q,Zhang Y D.Numerical simulation of segmented telescope[J].Acta Optica Sinica,2014,34(7):0722002.廖周,邱琪,张雨东.分块拼接望远镜的数值仿真[J].光学学报,2014,34(7):0722002.
[4]Zhang Y F.Research of image denoising and document image skew correction based on Shearlet[D].Kaifeng:Henan University,2016:5-20.张一凡.基于Shearlet变换的图像降噪及文档图像倾斜校正研究[D].开封:河南大学,2016:5-20.
[5]Boukharouba A.A new algorithm for skew correction and baseline detection based on the randomized Hough Transform[J].Journal of King Saud University-Computer and Information Sciences,2017,29(1):29-38.
[6]Ahmad R,Rashid S F,Afzal M Z,et al.A novel skew detection and correction approach for scanned documents[C]//DAS 2016,12th Intl IAPRWorkshop on Document Analysis Systems,Santorini,Greece,2016.
[7]Alaei A,Nagabhushan P,Pal U,et al.An efficient skew estimation technique for scanned documents:An application of piece-wise painting algorithm[J].Journal of Pattern Recognition Research,2016,11(1):1-14.
[8]O'Reilly C,Gluhak A,Imran M A.Distributed anomaly detection using minimum volume elliptical principal component analysis[J].IEEE Transactions on Knowledge and Data Engineering,2016,28(9):2320-2333.
[9]Verma R N,Malik L G.Review of illumination and skew correction techniques for scanned documents[J].Procedia Computer Science,2015,45(1):322-327.
[10]Easley G R,Labate D,Colonna F.Shearlet-based total variation for denoising[J].IEEE Transactions on Image Processing,2009,18(18):260-268.
[11]Jakobsen M,Lemvig J.Reproducing formulas for generalized translation invariant systems on locally compact abelian groups[J].Transactions of the American Mathematical Society,2016,368(12):8447-8480.
[12]Jenifha M A,Babu G M.Nonparametric illumination and skew correction for scanned document images[J].Digital Image Processing,2016,8(5):165-169.
[13]Hu J H.Image denoising based on Shearlet transform directionality[D].Xi′an:Northwest University,2014:14-26.胡江华.基于Shearlet变换方向性的图像消噪[D].西安:西北大学,2014:14-26.
[14]Koo H I,Cho N I.Robust skew estimation using straight lines in document images[J].Journal of Electronic Imaging,2016,25(3):033014.
[15]Wu Y Q,Tao F X.Multispectral and panchromatic image fusion based on improved projected gradient NMF in NSST domain[J].Acta Optica Sinica,2015,35(4):0410005.吴一全,陶飞翔.改进投影梯度NMF的NSST域多光谱与全色图像融合[J].光学学报,2015,35(4):0410005.
[16]Shi M H,Liu W.Anew image denoising algorithm based on shift-invariant Shearlet transform domain[J].Infrared Technology,2016,38(1):33-40.石满红,刘卫.一种新的平移不变Shearlet变换域图像去噪算法[J].红外技术,2016,38(1):33-40.
[17]Guo K,Labate D.Characterization and analysis of edges using the continuous Shearlet transform[J].Siam Journal on Imaging Sciences,2009,2(3):959-986.
[18]Easley G R,Labate D,Lim W Q.Optimally sparse image representations using shearlets[C]//Proceedings of IEEE International Conference on Signals,Systems and Computers,2006:974-978.
[19]Zhang J X.The theory of modern Chinese character pen shape[C].The Selection of the 2nd International Chinese Teaching Symposium.Beijing:Beijing Language Institute Press,1988:162-165.张静贤.现代汉字笔形论[C].第二届国际汉语教学讨论会论文选.北京:北京语言学院出版社,1988:162-165.
[20]Feng Z W.Computer processing of natural language[J].Chinese Information,1997(4):26-27.冯志伟.自然语言的计算机处理[J].中文信息,1997(4):26-27.
[21]Fan L.Research on stroke characteristics and feature of Chinese characters[D].Chongqing:Southwestern University,2013:17-27.范玲.现行汉字笔画分布及特征研究[D].重庆:西南大学,2013:17-27.
[22]Papandreou A,Gatos B,Louloudis G,et al.ICDAR2013document image skew estimation contest[C]//Proceedings of IEEE International Conference on Document Analysis and Recognition,2013:1444-1448.
[23]Chou C H,Chu S Y,Chang F.Estimation of skew angles for scanned documents based on piecewise covering by parallelograms[J].Pattern Recognition,2007,40(2):443-455.
[24]Dey P,Noushath S.e-PCP:A robust skew detection method for scanned document images[J].Pattern Recognition,2010,43(3):937-948.