Improving Ancient Roman Coin Classification by Fusing Exemplar-Based Classification and Legend Recognition

In this paper we present an image-based classification method for ancient Roman Republican coins that uses multiple sources of information. Exemplar-based classification, which estimates the coins' visual similarity by means of a dense correspondence field, and lexicon-based legend recognition are unified to a common classification approach. Classification scores from both coin sides are further integrated to an overall score determining the final classification decision. Experiments carried out on a dataset of 60 different classes comprising 464 coin images show that the combination of methods leads to higher classification rate than using them separately.

[1]  Matthieu Guillaumin,et al.  Segmentation Propagation in ImageNet , 2012, ECCV.

[2]  Martin Kampel,et al.  Coarse-to-Fine Correspondence Search for Classifying Ancient Coins , 2012, ACCV Workshops.

[3]  Laurens van der Maaten,et al.  COIN-O-MATIC: A fast system for reliable coin classification , 2006 .

[4]  Ognjen Arandjelovic,et al.  Automatic attribution of ancient Roman imperial coins , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[5]  Kai Wang,et al.  Word Spotting in the Wild , 2010, ECCV.

[6]  Igor Holländer,et al.  Dagobert - A New Coin Recognition and Sorting System , 2003, DICTA.

[7]  Václav Hlavác,et al.  Efficient MRF Deformation Model for Non-Rigid Image Matching , 2007, CVPR.

[8]  Minoru Fukumi,et al.  Rotation-invariant neural pattern recognition system with application to coin recognition , 1992, IEEE Trans. Neural Networks.

[9]  Matthijs C. Dorst.  Distinctive Image Features from Scale-Invariant Keypoints , 2011 .

[10]  Antonio Torralba,et al.  SIFT Flow: Dense Correspondence across Scenes and Its Applications , 2011, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Martin Kampel,et al.  Word detection applied to images of ancient Roman coins , 2012, 2012 18th International Conference on Virtual Systems and Multimedia.

[12]  Thomas Deselaers,et al.  ClassCut for Unsupervised Class Segmentation , 2010, ECCV.

[13]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[14]  Václav Hlavác,et al.  Efficient MRF Deformation Model for Non-Rigid Image Matching , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[15]  A. Chitra,et al.  A New Approach to Coin Recognition using Neural Pattern Analysis , 2005, 2005 Annual IEEE India Conference - Indicon.

[16]  Michael H. Crawford,et al.  Roman Republican coinage , 1975 .

[17]  Martin Kampel,et al.  Recognizing Ancient Coins Based on Local Features , 2008, ISVC.

[18]  C. J. Citron,et al.  Numismatics , 1933, The Classical Review.

[19]  Dana H. Ballard,et al.  Generalizing the Hough transform to detect arbitrary shapes , 1981, Pattern Recognit..

[20]  Daniel P. Huttenlocher,et al.  Pictorial Structures for Object Recognition , 2004, International Journal of Computer Vision.

[21]  Hans Burkhardt,et al.  An Efficient Gradient Based Registration Technique for Coin Recognition , 2006 .

[22]  Martin Kampel,et al.  Identification of ancient coins based on fusion of shape and local features , 2011, Machine Vision and Applications.

[23]  Ognjen Arandjelovic,et al.  Reading Ancient Coins: Automatically Identifying Denarii Using Obverse Legend Seeded Retrieval , 2012, ECCV.