Handedness detection of online handwriting based on horizontal strokes

Forensic Handwriting Experts (FHEs) have to expand their work field and elaborate supplementary methods because of spreading of online signatures. Special devices, as digitizing tablets, enable us to analyze additional information of online writings (for instance speed, pressure, x; y coordinates and related timestamp). These data are suitable for verification of expert's observations and may help in indentification and classification of groups of writers. This paper describes an automatized method based on approach of FHEs. Our special aim is to contribute to the scientific background of examination of handedness in Forensic Handwriting Examination. We attempted to ensure the efficiency of the investigation by using only significant features to classify the handwriting samples. Besides this, instead of a microscopic examination applied by FHEs, we evaluated absolutely objective data of online handwriting samples.