Detección de personas armadas en áreas públicas

Abstract

Abstract:With only 9% of the world's population, Latin America has one of the highest rates of violence in the world, generating insecurity, crime, robberies, increased weapons, and homicides. The phenomena generated by the possession of firearms are also linked to homicide cases in Ecuador, for 2010 to 2015 with a 70% presence in all period s firearms were the main objects used in these cases. In this degree project we work with object detection to create a system that can detect various types of weapons in public spaces such as stores, ATMs, streets, among others. For the creation of the system, the PyTorch API and TensorFlow were used; subsequently, 5 neural networks were trained and evaluated with the same libraries. The final results executed by the systems showed that YoloV3 with a 0.81% accuracy in detections is the most promising model to be implemented in real video security systems.