Segmentation and Classification of Pages for Digitized Documents of the Public Prosecutor’s Office

Rivera K.; Quintanilla D.; Espezua Á. · Communications in Computer and Information Science (2023) · DOI: 10.1007/978-3-031-35445-8_3

Resumen

The digitization of physical documents is a common objective in the process of digital transformation of public organizations in Peru, however this process has a particular task that consumes a lot of time and resources: the segmentation and classification of pages in digitized documents, that is, determine where one document ends and the next begins, as well as naming each classified document. To this end, in this article we present a solution that uses machine learning and regular expressions for the automatic segmentation of pages in digitized documents of the Public Prosecutor’s Office and their classification. Our solution achieved an accuracy of 93% using Random Forest for page segmentation and high accuracy in the classification of document types above 90% accuracy using regular expressions based on rules.

Instituciones: Universidad Nacional Mayor de San Marcos; Universidad Nacional Santiago Antúnez de Mayolo