摘要: This paper presents a new method of page segmentation based on the analysis background (white areas). The proposed is capable segmenting pages with non-rectangular layout as well various angles skew. characteristics are follows: (1) thinning enables us to represent white areas any shape connected thin lines or chains and robustness for tilted images also achieved by representation; (2) this representation, task defined find loops enclosing printed areas. eliminating unnecessary using not only feature areas, but black divided chain. Based experimental results comparison previous methods, we discuss advantages limitations method.
sci-hub.se 参考文章(7)
Theo Pavlidis, Jiangying Zhou, Page segmentation and classification Document image analysis. pp. 226- 238 ,(1995)
Haralick, Document image understanding: geometric and logical layout computer vision and pattern recognition. pp. 385- 390 ,(1994) , 10.1109/CVPR.1994.323855
H.S. Baird, S.E. Jones, S.J. Fortune, Image segmentation by shape-directed covers international conference on pattern recognition. pp. 820- 825 ,(1990) , 10.1109/ICPR.1990.118223
L. O'Gorman, The document spectrum for page layout analysis IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 15, pp. 1162- 1173 ,(1993) , 10.1109/34.244677
A. Antonacopoulos, R.T. Ritchings, Flexible page segmentation using the background international conference on pattern recognition. ,vol. 2, pp. 339- 344 ,(1994) , 10.1109/ICPR.1994.576932
Nicolas Normand, Christian Viard-Gaudin, None, A background based adaptive page segmentation algorithm international conference on document analysis and recognition. ,vol. 1, pp. 138- 141 ,(1995) , 10.1109/ICDAR.1995.598961
L. Lam, S.-W. Lee, C.Y. Suen, Thinning methodologies-a comprehensive survey IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 14, pp. 869- 885 ,(1992) , 10.1109/34.161346