Use this link to cite:
http://hdl.handle.net/2183/31196 Detección de intrusos a partir de imágenes submuestreadas
Loading...
Identifiers
Publication date
Authors
Fernández Graña, Javier
Advisors
Other responsabilities
Enxeñaría informática, Grao en
Journal Title
Bibliographic citation
Type of academic work
Academic degree
Abstract
[Resumen] En este proyecto se ha desarrollado un sistema de detección de intrusos en el cual las imágenes
capturadas se comprimen antes de transmitirse a un nodo central, que implementa una
estrategia de aprendizaje máquina para analizar los datos. La idea subyacente consiste en simular
la transmisión de un vídeo, el cual se trata como una secuencia de imágenes. Como el
proceso de detección exige altas capacidades de cómputo, esta secuencia se envía a un nodo
central de procesamiento. Una vez recibidos los datos, se ejecuta un algoritmo de detección
de intrusos construido con el framework OpenCV, mientras que el resto del sistema se ha simulado
usando Matlab. Por último, se realiza un estudio de los parámetros principales del
sistema de transmisión, y se razona como influyen estos en el rendimiento de la detección. Se
ha seguido una metodología iterativa incremental para el desarrollo del proyecto.
[Abstract] In this project, an intrusion detection system has been developed in which the captured images are compressed before being transmitted to a central node, which implements a machine learning strategy to analyze the data. The underlying idea is to simulate the transmission of a video, which is treated as a sequence of images. As the discovery process requires high computing capacities, this sequence is sent to a central processing node. Once the data is received, an intrusion detection algorithm built with the OpenCV framework is executed, while the rest of the system has been simulated using MatLab. Finally, a study of the main parameters of the transmission system is carried out, and it is explained how these influence the detection performance. An incremental iterative methodology has been followed for the development of the project.
[Abstract] In this project, an intrusion detection system has been developed in which the captured images are compressed before being transmitted to a central node, which implements a machine learning strategy to analyze the data. The underlying idea is to simulate the transmission of a video, which is treated as a sequence of images. As the discovery process requires high computing capacities, this sequence is sent to a central processing node. Once the data is received, an intrusion detection algorithm built with the OpenCV framework is executed, while the rest of the system has been simulated using MatLab. Finally, a study of the main parameters of the transmission system is carried out, and it is explained how these influence the detection performance. An incremental iterative methodology has been followed for the development of the project.
Description
Editor version
Rights
Atribución-NoComercial-SinDerivadas 3.0 España








