Character Recognition, Orientation, and Scale Estimation Thanks to the Fourier Mellin Transform

In this paper, we consider the general problem of technical document interpretation, applied to the documents of the French Telephonic Operator, France Telecom. More precisely, we focus the content of this paper on the computation of a new set of features allowing the classification of multioriented and multi-scaled patterns. This set of Invariant is based on the Fourier Mellin Transform. The interests of this computation rely on the possibility to use this Fourier Mellin transform within a "filtering mode", that permits to solve the well known difficult problem of connected character recognition. In this paper, we also present an original technique allowing to compute an estimation of the orientation of each shape to be recognized.