Invariants discretization for individuality representation in handwritten authorship

Abstract

Writer identification is one of the areas in pattern recognition that have created a center of attention by many researchers to work in. Its focal point is in forensics and biometric application as such the writing style can be used as biometric features for authenticating a writer. Handwriting style is a personal to individual and it is implicitly represented by unique features that are hidden in individual's handwriting. These unique features can be used to identify the handwritten authorship accordingly. Many researches have been done to develop algorithms for extracting good features that can reflect the authorship with good performance. However, this paper investigates the individuality representation of individual features through discretization technique. Discretization is a procedure to explore the partition of attributes into intervals and to unify the values for each interval. It illustrates the pattern of data systematically which improved the identification accuracy. An experiment has been conducted using IAM database with 3520 training data and 880 testing data (70% training data and 30% testing data) and 2639 training data and 1760 testing data (60% training data and 40% testing data). The results reveal that with invariants discretization, the accuracy of handwritten identification is improved significantly with the classification accuracy of 99.90% compared to undiscretized data.