Publication: Detección de manipulación en ficheros multimedia basado en algoritmos de compresión
Loading...
Files
Official URL
Full text at PDC
Publication Date
2018
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
Las imágenes y vídeos digitales juegan un papel muy importante en la vida cotidiana. A día de hoy, la mayor parte de la población es poseedora de cámaras fotográficas de última generación integradas en su dispositivo móvil. El desarrollo tecnológico no sólo facilita la generación de contenido multimedia, sino también la manipulación intencionada de éste, y es aquí donde las técnicas forenses de detección de manipulación sobre imágenes y vídeos cobran gran importancia. En este trabajo se proponen dos metodologías forenses basadas en algoritmos de compresión: La primera de ellas trata de detectar la presencia de recompresión en un vídeo digital mediante el análisis de sus macrobloques, característica propia del estándar H.264-MPEG4. Posteriormente, se utiliza la máquina de soporte vectorial para crear el modelo que permita la verificación del número de recompresiones de un vídeo. La segunda metodología que se explica en este trabajo tiene por objetivo detectar alteraciones de tipo ‘empalme’, es decir, regiones que no pertenecen al contenido original de una imagen digital, técnica que está basada en la tasa de error que introduce el algoritmo de compresión JPEG cada vez que recomprime una imagen.
Digital images and videos play an important role in everyday life. Today, the majority of the population owns state-of-the-art cameras integrated into their mobile device. Technological development not only facilitates the generation of multimedia content, but also the intentional manipulation of it, and this is where forensic techniques of detecting manipulation on images and videos take on great importance. In this work two forensic methodologies based on compression algorithms are proposed: The first one tries to detect the presence of recompression in a digital video by means of the analysis of its macroblocks, characteristic of the H.264-MPEG4 standard. Subsequently, the vectorial support machine is used to create the model that allows the verification of the number of recompressions of a video. The second methodology explained in this work aims to detect splicing alterations, i.e. regions that do not belong to the original content of a digital image, a technique based on the error rate introduced by the JPEG compression algorithm each time it recompresses an image.
Digital images and videos play an important role in everyday life. Today, the majority of the population owns state-of-the-art cameras integrated into their mobile device. Technological development not only facilitates the generation of multimedia content, but also the intentional manipulation of it, and this is where forensic techniques of detecting manipulation on images and videos take on great importance. In this work two forensic methodologies based on compression algorithms are proposed: The first one tries to detect the presence of recompression in a digital video by means of the analysis of its macroblocks, characteristic of the H.264-MPEG4 standard. Subsequently, the vectorial support machine is used to create the model that allows the verification of the number of recompressions of a video. The second methodology explained in this work aims to detect splicing alterations, i.e. regions that do not belong to the original content of a digital image, a technique based on the error rate introduced by the JPEG compression algorithm each time it recompresses an image.
Description
Trabajo de Fin de Grado, Universidad Complutense, Facultad de Informática, Departamento de Ingeniería del Software e Inteligencia Artificial, Curso 2017/2018