Desarrollo de un sistema de visión artificial para la detección de accidente y/o congestionamiento vehicular.

Currently the world is full of devices capable of to capture images over long distances, monitor specific sectors, security cameras for businesses, eagle eyes, etc. All these devices are very useful when making a decision by humans, but this cannot be present 24 hours, 365 days a year, working and making decisions without taking a break, this is where the machine vision plays a role key, making such devices become passive devices capture images of assets for themselves, able to make a decision as you would a thinking, intelligent human being, regardless of the schedule. Nowadays, it is necessary to determine the vehicle capacity of a particular place where there is often congestion of vehicles and that traffic is controlled by traffic lights with timers that are at intersections to regulate the normal vehicular and pedestrian traffic, but are inefficient when you agglomerations peaking occur, causing a series of problems of environmental pollution, traffic accidents, noise pollution, etc. Under this scenario, what is presented is a proposal framed within the project of intelligent traffic lights; revealing certain tools and algorithms that can help in the detection of possible congestion and traffic accidents autonomously, without the need for the competent authority is present. This degree work includes analysis, design and implementation of a system based on artificial vision for detecting congestion and traffic accidents from the perspective of a traffic light system.

• 1. Verfasser: Aponte Rueda, Pedro Fernando,Andrés Armando Armijos Armijos (author)
• Format: bachelorThesis
• Sprache: spa
• Veröffentlicht: 2016
• Schlagworte: <VISIÓN ARTIFICIAL> <DETECCIÓN DE VEHÍCULOS> <UBUNTU 14.04> <LENGUAJE C++> <FRAMEWORK QT> <PROCESAMIENTO DIGITAL DE IMÁGENES> <CLASIFICADORES EN CASCADA> <OPEN CV>
• Online Zugang: http://dspace.unl.edu.ec/jspui/handle/123456789/11588