Segmentation and Retrieval of Ancient Graphic Documents

Uttama, Surapong; Loonis, Pierre; Delalandre, Matthieu; Ogier, Jean-Marc

doi:10.1007/11767978_8

Cited by 16 publications

(16 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With template matching based approaches the similarity among images is computed by a simple value comparison [11] or using a specific dissimilarity function [12]. In Tan et al [13], the similarity is computed by means of the Chi-square distance using the distribution of tf-idf vectors, while Bhattacharyya distance between two histograms is used in [14]. Fataicha et al [15] proposed a measure based on minimum edit-distance.…”

Section: A Similarity Measuresmentioning

confidence: 99%

Unsupervised Classification of Structurally Similar Document Images

Kumar

Doermann

2013

2013 12th International Conference on Document Analysis and Recognition

View full text Add to dashboard Cite

In this paper, we present a learning based approach for computing structural similarities among document images for unsupervised exploration in large document collections. The approach is based on multiple levels of content and structure. At a local level, a bag-of-visual words based on SURF features provides an effective way of computing content similarity. The document is then recursively partitioned and a histogram of codewords is computed for each partition. Structural similarity is computed using a random forest classifier trained with these histogram features. We experiment with three diverse datasets of document images varying in size, degree of structural similarity, and types of document images. Our results demonstrate that the proposed approach provides an effective general framework for grouping structurally similar document images. I. INTRODUCTIONAssociating images with similar structural characteristics is a first and important step in mining visual information from large collections of images. Having an effective image similarity measure makes searching and browsing large repositories of images easier as clustered images can be collectively tagged with attributes and indexed for search and retrieval tasks. Clustering search results for queries on document image collections, or performing near-duplicate detection are examples of applications where good similarity measures are needed.

show abstract

Section: A Similarity Measuresmentioning

confidence: 99%

Unsupervised Classification of Structurally Similar Document Images

Kumar

Doermann

2013

2013 12th International Conference on Document Analysis and Recognition

View full text Add to dashboard Cite

show abstract

“…Several studies (Uttama et al, 2006;Journet et al, 2008) proposed to characterize and index images of ancient documents by their content by exploring textural analysis based on statistical and spectral properties of texture. The authors of (Uttama et al, 2006) introduce a segmentation method of drop caps based on a combination of different texture analysis approaches such as the GLCM (Haralick et al, 1973) and the autocorrelation function (Petrou and Sevilla, 2006). In (Journet et al, 2008), the authors propose an extraction algorithm of texture features that is devoted to the analysis of historical documents.…”

Section: Texture Feature Extractionmentioning

confidence: 99%

“…The extracted texture features are mainly investigated and analyzed separately in independent experiments for document analysis (Journet et al, 2008). Some works deal with the whole ancient document image (Journet et al, 2008) and others are applied to graphic images such as drop caps (Uttama et al, 2006;Coustaty et al, 2011). There have been few comparative studies on gradient, multiple channel Gabor filters, and co-occurrence features (Payne et al, 1994;Said et al, 2000;Liu et al, 2005;Ding et al, 2007;Zhu et al, 2001) for document segmentation, character recognition, and script and language identification.…”

Section: Introductionmentioning

confidence: 99%

A Pixel Labeling Framework for Comparing Texture Features Application to Digitized Ancient Books

Mehri

Gomez‐Krämer²,

Héroux

et al. 2014

Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods

View full text Add to dashboard Cite

“…However, the main interest of this study is based on specific graphics called drop caps, and on the extraction of shapes in drop caps and particularly on the most important shape : the letter itself. This work is inspired by [PV06] and [ULDO05] which used a Zipf law and a Wold decomposition to extract elements of drop caps.…”

Section: Navidomassmentioning

confidence: 99%