Hermite and Gabor transforms for noise reduction and handwriting classification in ancient manuscripts

@article{citeulike:1851600, abstract = {Abstract\ \ In this paper, we propose a biologically inspired, global and segmentation free methodology for manuscript noise reduction and classification. Our method consists of developing well-adapted tools for writing enhancement, background noise, text and drawing separation and handwritten patterns characterization with orientation features. We have used here analysis of handwritten images in the spectral domain by frequency decompositions (Hermite transforms) and Gabor filtering for selective text information extraction. We have tested our approach of writing classification on ancient manuscripts corpus, mainly composed of 18th century authors’ documents. The current results are very promising: they show that our biologically inspired methodology can be efficiently used for handwriting analysis without any a priori grapheme segmentation.}, author = {Eglin, V\'{e}ronique and Bres, St\'{e}phane and Rivero, Carlos }, citeulike-article-id = {1851600}, doi = {http://dx.doi.org/10.1007/s10032-007-0039-z}, journal = {International Journal on Document Analysis and Recognition}, keywords = {classifier, historical, image-processing, interlending, preprocessing}, month = {April}, number = {2}, pages = {101--122}, posted-at = {2007-11-01 16:04:14}, priority = {5}, title = {Hermite and Gabor transforms for noise reduction and handwriting classification in ancient manuscripts}, url = {http://dx.doi.org/10.1007/s10032-007-0039-z}, volume = {9}, year = {2007} }

TY - JOUR ID - citeulike:1851600 L3 - citeulike-article-id:1851600 TI - Hermite and Gabor transforms for noise reduction and handwriting classification in ancient manuscripts JF - International Journal on Document Analysis and Recognition VL - 9 IS - 2 SP - 101 EP - 122 N2 - Abstract  In this paper, we propose a biologically inspired, global and segmentation free methodology for manuscript noise reduction and classification. Our method consists of developing well-adapted tools for writing enhancement, background noise, text and drawing separation and handwritten patterns characterization with orientation features. We have used here analysis of handwritten images in the spectral domain by frequency decompositions (Hermite transforms) and Gabor filtering for selective text information extraction. We have tested our approach of writing classification on ancient manuscripts corpus, mainly composed of 18th century authors’ documents. The current results are very promising: they show that our biologically inspired methodology can be efficiently used for handwriting analysis without any a priori grapheme segmentation. KW - classifier KW - historical KW - image-processing KW - interlending KW - preprocessing AU - Eglin, Véronique AU - Bres, Stéphane AU - Rivero, Carlos PY - 2007/04/05/ UR - http://dx.doi.org/10.1007/s10032-007-0039-z ER -