Laser Printer Attribution: Exploring new Features and Beyond Dataset

This is the dataset of the paper Laser Printer Attribution: Exploring New Features and Beyond.

The dataset contains printed documents and extracted characters from 10 printers as described in the
paper. Each printer prints documents in English and Portuguese. The directories are then divided in
english documents (I) and portuguese Documents (P). There are three kinds of dataset:

1- Printer_full_dataset: this dataset contains the full documents scanned. We use this dataset when
we perform experiments on frames and in the whole document.

2- Printer_char_dataset: contains extracted characters per printer and per document. We use this
dataset when we perform experiments on characters.

3- Printer_char_Farid_Dataset: The work of kee and Farid was applied on letters of same
size, and this does not happen on our Printer_char_dataset. Here, the letters are resized to be of
same dimensions, then, the Kee and Farid's approach can work.

Please find more details on the dataset in the README file that follows the zip file.

Once using the dataset in your research, plese don't forget to cite our paper:

@article{anselmoetal:2015,
title = "Laser Printer Attribution: Exploring New Features and Beyond",
journal = "Forensic Science International ",
volume = "247",
number = "0",
pages = "105 - 125",
year = "2015",
author = "Anselmo Ferreira and Luiz C. Navarro and Giuliano Pinheiro and Jefersson A. dos Santos and Anderson Rocha",
}