Feature extraction from handwritten documents for personality analysis

Handwriting can be used to predict or analyze a person's behavioral or personality traits. Characteristics of a handwriting are studied for that. In this work various characteristics like size, spacing, slant, skew, pressure, etc are studied. Since signature reflects important characteristics of human being, we analyze them also. As our objective is to build an automated or computerized handwriting analysis system, we discuss the features from algorithmic view points. For experimental purpose, various handwriting of different writers are collected. Identical as well as different texts is used for that. Implementing these algorithms on the data sets, we obtain experimental results which are given here.