ICDAR 2021 Competition on Historical Map Segmentation.
作者： Edwin Carlinet , Thierry Géraud , Ladislav Lenc , Clément Mallet , Joseph Chazalon
DOI:
关键词:
摘要: This paper presents the final results of ICDAR 2021 Competition on Historical Map Segmentation (MapSeg), encouraging research a series historical atlases Paris, France, drawn at 1/5000 scale between 1894 and 1937. The competition featured three tasks, awarded separately. Task~1 consists in detecting building blocks was won by L3IRIS team using DenseNet-121 network trained weakly supervised fashion. task is evaluated 3 large images containing hundreds shapes to detect. Task~2 segmenting map content from larger sheet, UWB U-Net-like FCN combined with binarization method increase detection edge accuracy. Task~3 locating intersection points geo-referencing lines, also who used dedicated pipeline combining binarization, line Hough transform, candidate filtering, template matching for refinement. Tasks~2 and~3 are 95 sheets complex content. Dataset, evaluation tools available under permissive licensing \url{https://icdar21-mapseg.github.io/}.
arxiv.org参考文章(12)
Philipp Fischer, Thomas Brox, None, U-Net: Convolutional Networks for Biomedical Image Segmentation medical image computing and computer assisted intervention. pp. 234- 241 ,(2015) , 10.1007/978-3-319-24574-4_28
Aurélie Lemaitre, Jean Camillerapp, Bertrand Coüasnon, Multiresolution cooperation makes easier document structure recognition International Journal on Document Analysis and Recognition. ,vol. 11, pp. 97- 109 ,(2008) , 10.1007/S10032-008-0072-6
I. Leplumey, J. Camillerapp, C. Queguiner, Kalman filter contributions towards document segmentation international conference on document analysis and recognition. ,vol. 2, pp. 765- 769 ,(1995) , 10.1109/ICDAR.1995.602015
Oliver Nina, Bryan Morse, William Barrett, A recursive Otsu thresholding method for scanned document binarization workshop on applications of computer vision. pp. 307- 314 ,(2011) , 10.1109/WACV.2011.5711519
Jorge Hernández, Beatriz Marcotegui, None, Morphological segmentation of building façade images international conference on image processing. pp. 3977- 3980 ,(2009) , 10.1109/ICIP.2009.5413756
Yves Grandvalet, Yoshua Bengio, Semi-supervised Learning by Entropy Minimization neural information processing systems. ,vol. 17, pp. 529- 536 ,(2004)
Marcus D Bloice, Peter M Roth, Andreas Holzinger, Biomedical image augmentation using Augmentor. Bioinformatics. ,vol. 35, pp. 4522- 4524 ,(2019) , 10.1093/BIOINFORMATICS/BTZ259
Gao Huang, Zhuang Liu, Geoff Pleiss, Laurens Van Der Maaten, Kilian Weinberger, Densely Connected Convolutional Networks computer vision and pattern recognition. pp. 2261- 2269 ,(2017) , 10.1109/CVPR.2017.243
Alexander Kirillov, Kaiming He, Ross Girshick, Carsten Rother, Piotr Dollar, Panoptic Segmentation 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). ,(2019) , 10.1109/CVPR.2019.00963
Jean Christophe Burie Van Nguyen, Christophe Rigaud, Arnaud Revel, A learning approach with incomplete pixel-level labels for deep neural networks. Neural Networks. ,vol. 130, pp. 111- 125 ,(2020) , 10.1016/J.NEUNET.2020.06.025