2014 - 2021
Criminal activities vary in scope and complexity, but exist in all sectors of society. With technological development, crimes have become more sophisticated, migrating from the physical world and also reaching the virtual world. Given the countless forms that illegal activities can take, the solution for modern society is not to invest in improved and scientifically based investigation techniques. With such sophisticated and increasingly complex demands comes the obligation to increasingly strengthen the foundations on which Forensic Science develops. More than ever, we need objective and scientifically based criteria to identify characteristics at crime and accident sites, or even to prevent the occurrence of crimes by taking effective preventive electronic surveillance actions. We need innovative approaches and solutions to current challenges to help us resolve the three basic questions regarding an illegal activity or large-scale accident: “Who?”, “Under what circumstances?” it is because?. Faced with so many challenges, this CAPES Pró-Forenses N 25/2014 project aims to develop algorithmic visual computing and machine intelligence solutions for problems related to forensic computing, digital security and electronic surveillance. The problems of interest are: (P1) detection of forgeries in digital images and videos; (P2) assignment of data capture source such as camera, scanner, printer; (P3) detection of clandestine plantations of, for example, Cannabis sativa from remote sensing images; (P4) development of human identification techniques based on faces; (P5) vehicle license plate recognition from digital videos; and (P6) analysis of activities in digital videos. The research will be carried out jointly with several partners at universities in Brazil and around the world, and with the Federal Police of Brazil, more specifically with the Audiovisual and Electronic Expertise Service (SEPAEL) of the National Institute of Criminalistics (INC).