PROBLEM TO BE SOLVED: To accurately perform writer identification by suppressing personal change included in analysis target handwriting.;SOLUTION: A feature point coordinate measurement section 202 extracts feature point coordinates of characters respectively included in analysis target handwriting data 21 and reference handwriting data 22 which is handwriting data of a comparison subject. A geometric normalization section 203 normalizes the handwriting data so that a coordinate difference between the handwriting data and a reference pattern at each feature point becomes the smallest for the entire character. A personal change suppression processing section 204 performs projective transformation for the normalized handwriting data so that a difference between coordinates of the plurality of pieces of normalized handwriting data and average coordinates of the handwriting at each feature point becomes the smallest for the entire character. A writer identification section 207 performs writer identification on the basis of the difference between the corresponding feature point coordinates of the projective-transformed analysis target handwriting data and the projective-transformed reference handwriting data.;COPYRIGHT: (C)2013,JPO&INPIT
展开▼