Sistema de identificação entomológica forense: uma abordagem usando LabVIEW

Abstract

Forensic entomology is one of the most accurate resources available for identifying origin of certain crimes, such as homicides, spawning grounds, drug trafficking, and smuggling. However, the lack of a complete police entomological database and an automatic recognition system to identify this type of traces have delayed the solution of countless crimes of high social repercus­sion. Forensic Engineering group Edmund Loccard of UTFPR, convened to the Department of Federal Police and CAPES developed a prototype of a regional entomological database, where methods of image pattern recognition are applied. This work aimed to develop and test an automatic identification system for these insect species, specifically using the LabVIEW tool. Several algorithms were implemented in this application, where images of partial fragments of insects, collected as traces at crime sites, were compared to the database. The ultimate goal is to identify the origin of the drug seized and the consequent identification of suspicious groups. The results showed that it was possible to establish a matching between the base of images trained with test images as well as images containing only fragments of insects.