{"type":"ScholarlyArticle","title":"Multi-task Layout Analysis for Historical Handwritten Documents Using Fully Convolutional Networks","authors":[{"name":"Yue Xu","slug":"yue-xu","given":"Yue","family":"Xu","affiliation":"National Laboratory of Pattern Recognition , Institute of Automation of Chinese Academy of Sciences , 95 Zhongguan East Road , Beijing 100190 , P.R. China; University of Chinese Academy of Sciences , Beijing , P.R. China"},{"name":"Fei Yin","slug":"fei-yin","given":"Fei","family":"Yin","affiliation":"National Laboratory of Pattern Recognition , Institute of Automation of Chinese Academy of Sciences , 95 Zhongguan East Road , Beijing 100190 , P.R. China; University of Chinese Academy of Sciences , Beijing , P.R. China"},{"name":"Zhaoxiang Zhang","slug":"zhaoxiang-zhang","given":"Zhaoxiang","family":"Zhang","affiliation":"National Laboratory of Pattern Recognition , Institute of Automation of Chinese Academy of Sciences , 95 Zhongguan East Road , Beijing 100190 , P.R. China; University of Chinese Academy of Sciences , Beijing , P.R. China; CAS Center for Excellence of Brain Science and Intelligence Technology , Beijing , P.R. China"},{"name":"Cheng-Lin Liu","slug":"cheng-lin-liu","given":"Cheng-Lin","family":"Liu","affiliation":"National Laboratory of Pattern Recognition , Institute of Automation of Chinese Academy of Sciences , 95 Zhongguan East Road , Beijing 100190 , P.R. China; University of Chinese Academy of Sciences , Beijing , P.R. China; CAS Center for Excellence of Brain Science and Intelligence Technology , Beijing , P.R. China"}],"abstract":"Layout analysis is a fundamental process in document image analysis and understanding. It consists of several sub-processes such as page segmentation, text line segmentation, baseline detection and so on. In this work, we propose a multi-task layout analysis method that use a single FCN model to solve the above three problems simultaneously. The FCN is trained to segment the document image into different regions and detect the center line of each text line by classifying pixels into different categories. By supervised learning on document images with pixel-wise labels, the FCN can extract discriminative features and perform pixel-wise classification accurately. After pixel-wise classification, post-processing steps are taken to reduce noises, correct wrong segmentations and find out overlapping regions. Experimental results on the public dataset DIVA-HisDB containing challenging medieval manuscripts demonstrate the effectiveness and superiority of the proposed method.","doi":"10.24963/ijcai.2018/147","url":"https://ijcai.org/proceedings/2018/147","pdf":"https://ijcai.org/proceedings/2018/0147.pdf","markdown":"https://ijcai.org/proceedings/2018/147.md","bibtex":"https://ijcai.org/proceedings/2018/bibtex/147","volume":{"slug":"2018","title":"Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2018","url":"https://ijcai.org/proceedings/2018/","year":2018,"location":"Stockholm","dates":"13-19 July 2018","editors":"Jérôme Lang","isbn":"978-0-9992411-2-7","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2018"},"section":"Computer Vision","track":"Main track","number":147,"pages":{"first":1057,"last":1063},"keywords":["Computer Vision: Language and Vision","Computer Vision: Computer Vision"],"videos":[],"references":[{"position":1,"raw":"[Al-Badr and Mahmoud, 1995] Badr Al-Badr and Sabri Mahmoud. Mahmoud, s.a.: Survey and bibliography of arabic optical text recognition. In Proceedings of the Sig- nal Processing, pages 49-77, 1995.","title":"Badr Al-Badr and Sabri Mahmoud. Mahmoud, s.a.: Survey and bibliography of arabic optical text recognition","authors":"Al-Badr, Mahmoud","year":"1995","venue":"Proceedings of the Signal Processing","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bukhari et al., 2012] Syed Saqib Bukhari, Thomas M. Breuel, Abedelkadir Asi, and Jihad El-Sana. Layout anal- ysis for arabic historical document images using machine learning. In Proceedings of the 13th International Con- ference on Frontiers in Handwriting Recognition, pages 639-644, 2012.","title":"Layout analysis for arabic historical document images using machine learning","authors":"Saqib Syed, Thomas M Bukhari, Abedelkadir Breuel, Jihad Asi, El-Sana","year":"2012","venue":"Proceedings of the 13th International Conference on Frontiers in Handwriting Recognition","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Chakraborty and Pal, 2016] Dibyayan Chakraborty and Umapada Pal. Baseline detection of multi-lingual un- constrained handwritten text lines. In Proceedings of the Pattern Recognition Letters, pages 74-81, 2016.","title":"Dibyayan Chakraborty and Umapada Pal. Baseline detection of multi-lingual unconstrained handwritten text lines","authors":"Pal Chakraborty","year":"2016","venue":"Proceedings of the Pattern Recognition Letters","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Chen and Wu, 2009] Yenlin Chen and Bingfei Wu. A multi- plane approach for text segmentation of complex docu- ment images. In Proceedings of the Pattern Recognition, pages 1419-1444, 2009.","title":"A multiplane approach for text segmentation of complex document images","authors":"Chen, Wu, Yenlin Chen, Bingfei Wu","year":"2009","venue":"Proceedings of the Pattern Recognition","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Faisal et al., 2005] Farooq Faisal, Govindaraju Venu, and Perrone Michael. Pre-processing methods for handwrit- ten arabic documents. In Proceedings of the International Conference on Document Analysis and Recognition, pages 267-271, 2005.","title":"Pre-processing methods for handwritten arabic documents","authors":"Farooq Faisal, Govindaraju Venu, Perrone Michael","year":"2005","venue":"Proceedings of the International Conference on Document Analysis and Recognition","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Gruning et al., 2017] Tobias Gruning, Roger Labah- n, Markus Diem, Florian Kleber, and Stefan Fiel. Read-bad: A new dataset and evaluation scheme for baseline detection in archival documents. In CoRR, page abs/1705.03311, 2017.","title":"Read-bad: A new dataset and evaluation scheme for baseline detection in archival documents","authors":"Gruning","year":"2017","venue":"CoRR","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Jia et al., 2014] Yangqing Jia, Evan Shelhamer, Jeff Don- ahue, Sergey Karayev, Jonathan Long, Ross Girshick, Ser- gio Guadarrama, and Trevor Darrell. Caffe: Convolutional architecture for fast feature embedding. In Proceedings of the 22nd ACM International Conference on Multimedia, pages 675-678, 2014.","title":"Caffe: Convolutional architecture for fast feature embedding","authors":"Jia","year":"2014","venue":"Proceedings of the 22nd ACM International Conference on Multimedia","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Li et al., 2008] Yi Li, Yefeng Zheng, David Doermann, and Stefan Jaeger. Script-independent text line segmentation in freestyle handwritten documents. In Proceedings of the Pattern Analysis and Machine Intelligence, pages 1313- 1329, 2008.","title":"Script-independent text line segmentation in freestyle handwritten documents","authors":"Yi Li, Yefeng Zheng, David Doermann, Stefan Jaeger","year":"2008","venue":"Proceedings of the Pattern Analysis and Machine Intelligence","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Long et al., 2015] J. Long, E. Shelhamer, and T. Darrell. Fully convolutional networks for semantic segmentation. In Proceedings of the International Conference on Com- puter Vision and Pattern Recognition, page 3431-3440, 2015.","title":"Fully convolutional networks for semantic segmentation","authors":"J Long, E Shelhamer, T Darrell","year":"2015","venue":"Proceedings of the International Conference on Computer Vision and Pattern Recognition","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Mehri et al., 2015] Maroua Mehri, Nibal Nayef, Pierre Héroux, Petra Gomez-Krämer, and Rémy Mullot. Learn- ing texture features for enhancement and segmentation of historical document images. In Proceedings of the 3rd International Workshop on Historical Document Imaging and Processing, pages 47-54, 2015.","title":"Learning texture features for enhancement and segmentation of historical document images","authors":"Mehri","year":"2015","venue":"Proceedings of the 3rd International Workshop on Historical Document Imaging and Processing","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Nagy et al., 1992] George Nagy, Mahesh Viswanathan, and Sharad Seth. A prototype document image analysis system for technical journals. In Proceedings of the IEEE, pages 10-22, 1992. [Ouwayed and Belaïd, 2008] Nazih Ouwayed and Abdel Belaïd. Multi-oriented text line extraction from handwrit- ten arabic documents. In Proceedings of the Eighth Iapr International Workshop on Document Analysis Systems, pages 339-346, 2008.","title":"A prototype document image analysis system for technical journals","authors":"George Nagy, Mahesh Viswanathan, Sharad Seth","year":"1992","venue":"Proceedings of the Eighth Iapr International Workshop on Document Analysis Systems","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Pechwitz and Margner, 2002] M. Pechwitz and V. Margner. Baseline estimation for arabic handwritten words. In Pro- ceedings Eighth International Workshop on Frontiers in Handwriting Recognition, pages 479-484, 2002.","title":"Baseline estimation for arabic handwritten words","authors":"M Pechwitz, V Margner","year":"2002","venue":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Razak et al., 2008] Zaidi Razak, Khansa Zulkiflee, Mo- hd Idris, Emran Mohd Tamil, Mohd Noorzaily Mo- hamed Noor, Rosli Salleh, Mohd Yaakob, Zulkifli Mo- hd Yusof, and Mashkuri Yaacob. Off-line handwriting tex- t line segmentation: A review. In International Journal of Computer Science and Network Security, pages 12-20, 2008.","title":"Off-line handwriting text line segmentation: A review","authors":"Khansa Zaidi Razak, Mohd Zulkiflee, Emran Idris, Mohd Mohd Tamil, Mohamed Noorzaily, Rosli Noor, Mohd Salleh, Zulkifli Yaakob, Mashkuri Mohd Yusof, Yaacob","year":"2008","venue":"International Journal of Computer Science and Network Security","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Simistira et al., 2016] Foteini Simistira, Mathias Seuret, Nicole Eichenberger, Angelika Garz, Marcus Liwicki, and Rolf Ingold. Diva-hisdb: A precisely annotated large dataset of challenging medieval manuscripts. In Proceed- ings of the 15th International Conference on Frontiers in Handwriting Recognition, pages 471-476, 2016.","title":"Diva-hisdb: A precisely annotated large dataset of challenging medieval manuscripts","authors":"Simistira","year":"2016","venue":"Proceedings of the 15th International Conference on Frontiers in Handwriting Recognition","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Simistira et al., 2017] Fotini Simistira, Manuel Bouillon, and Mathias Seuret. Icdar2017 competition on layout anal- ysis for challenging medieval manuscripts. In Proceedings of the 14th International Conference on Document Analy- sis and Recognition, pages 1361-1370, 2017.","title":"Icdar2017 competition on layout analysis for challenging medieval manuscripts","authors":"Fotini Simistira, Manuel Bouillon, Mathias Seuret","year":"2017","venue":"Proceedings of the 14th International Conference on Document Analysis and Recognition","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Simonyan and Zisserman, 2015] Karen Simonyan and An- drew Zisserman. Very deep convolutional networks for large-scale image recognition. In Proceedings of the Inter- national Conference on Learning Representations, 2015.","title":"Very deep convolutional networks for large-scale image recognition","authors":"Karen Simonyan, Andrew Zisserman","year":"2015","venue":"Proceedings of the International Conference on Learning Representations","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Uttama et al., 2005] Surapong Uttama, Jean Marc Ogier, and Pierre Loonis. Top-down segmentation of ancien- t graphical drop caps: lettrines. In Proceedings of the International Workshop on Graphics Recognition, pages 87-96, 2005.","title":"Top-down segmentation of ancient graphical drop caps: lettrines","authors":"Uttama","year":"2005","venue":"Proceedings of the International Workshop on Graphics Recognition","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Xavier and Yoshua, 2010] Glorot Xavier and Bengio Yoshua. Understanding the difficulty of training deep feedforward neural networks. In Proceedings of the 13th International Conference on Artificial Intelligence and Statistics, pages 249-256, 2010.","title":"Understanding the difficulty of training deep feedforward neural networks","authors":"Yoshua ; Xavier, Glorot Xavier, Bengio Yoshua","year":"2010","venue":"Proceedings of the 13th International Conference on Artificial Intelligence and Statistics","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Xie and Tu, 2015] Saining Xie and Zhuowen Tu. Holistically-nested edge detection. In Proceedings of the 15th International Conference on Computer Vision, pages 1395-1403, 2015.","title":"Holistically-nested edge detection","authors":"Tu ; Xie, Saining Xie, Zhuowen Tu","year":"2015","venue":"Proceedings of the 15th International Conference on Computer Vision","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Yin and Liu, 2009] Fei Yin and Cheng-Lin Liu. Handwrit- ten chinese text line segmentation by clustering with dis- tance metric learning. In Pattern Recognition, pages 3146- 3157, 2009.","title":"Handwritten chinese text line segmentation by clustering with distance metric learning","authors":"Fei Yin, Cheng-Lin Liu","year":"2009","venue":"Pattern Recognition","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}