Application of the gsp_m algorithm for the identification of behavioral patterns of people who shoplift
| Autor: | Hector F Gomez A, Victor H Cordova A, E. Freddy Robalino P, Alfredo Vinicio Zuniga Tinizaray, Carlos Eduardo Martinez C, Boris Marcel Diaz Pauta, Edwin Fabricio Lozada T |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science
Behavioral pattern 02 engineering and technology 01 natural sciences Domain (software engineering) Set (abstract data type) Identification (information) Component (UML) 0103 physical sciences 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Suspect 010306 general physics Algorithm |
| Zdroj: | ICNC-FSKD |
| DOI: | 10.1109/fskd.2018.8687200 |
| Popis: | This work analyzes the results of the use of the algorithm GSP_M (Generalized Sequential Patterns + memory)for processing videos of surveillance in supermarkets, in order to identify the perpetration of thefts. It describes the theoretical framework of concepts used in video surveillance systems, their architecture, and, characteristics. It justified the use of sequential pattern identification algorithms as a tool to analyze human behavior in domains of shoplifting and mathematically define some concepts used by the tool. It describes the objectives of the application of video surveillance systems in supermarkets to minimize shoplifting, explaining that this is based on the fact that, monitored people have a behavior that can be characterized by a set of sequential actions, whose analysis allows obtaining conclusions to consider a person as a suspect, generating an alert or definitely an alarm. The GSP and GSP_M algorithms are described and compared. It details the procedure defined to evaluate the performance of the algorithm GSP_M in the domain of shoplifting, and, the comparison of the results obtained with the results of the GSP application. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|