Distance Based Strategy for Supervised Document Image Classification

Fabien Carmagnac (1, 2) , Pierre Héroux (2) , Eric Trupin (2)

1 A2iA SA.
2 PSI - Perception, Systèmes, Information

This paper deals with supervised document image classification. An original distance based strategy allows automatic feature selection. The computation of a distance between an image to be classified and a class representative (point of view) allows to estimate a membership function for all classes. The choice of the best point of view performs the feature selection. This idea is used by an algorithm which iteratively filters the list of candidate classes. The training phase is performed by computing the distances between every class. Each iteration of the classification algorithm computes the distance d between the image to be classified and the chosen representative. The classes whose distance with this point of view differs from d are deleted in the list of candidate classes. This strategy is implemented as a module of A2IA FieldReader to identify the class of the processed document. Experimental results are presented and compared with results given by a knn classifier.

Domaines: Traitement du texte et du document

Fabien Carmagnac, Pierre Héroux, Eric Trupin. Distance Based Strategy for Supervised Document Image Classification. IAPR International joint workshops on Statistical Pattern Recognition and Syntactic and Structural Pattern Recognition, 2004, France. pp.894-902, ⟨10.1007/978-3-540-27868-9_98⟩. ⟨hal-00605643⟩