A MULTISCALE AND MULTIORIENTATION RECOGNITION TECHNIQUE APPLIED TO DOCUMENT INTERPRETATION: APPLICATION TO THE FRENCH TELEPHONE NETWORK MAPS

摘要

In this paper, we propose an original methodology which allows the detection and the recognition of multioriented and multiscaled patterns. The supports on which the method is applied are technical documents representing the network of the French Telephone operator France Telecom, overlaid on urban maps. The adopted technique, based on the Fourier-Mellin transform (FMT) is integrated in a global strategy that solves ambiguous situations, through the providing of contextual information. The strategy, which is applied to solve the character and symbol classification problem, can be divided into two stages. The first one consists in constructing a set of invariant descriptors for each isolated pattern which is extracted from a characters layer given by the system approach. The second stage, based on a filtering scheme, consists in detecting and recognizing connected shapes. The results of the application of this technique are very encouraging since the classification rate reaches excellent scores in comparison with classical techniques.