Implementación y análisis de un algoritmo de compresión de imágenes basado en JPEG-XL
Loading...
Official URL
Full text at PDC
Publication date
2024
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citation
Abstract
Una imagen digital es una representación en dos dimensiones de una imagen en una matriz, que es almacenada y reproducida por un ordenador u otro dispositivo digital. Como ocurre con cualquier objeto que es guardado en alguna parte, ocupa un espacio. En este caso, no se trata de un espacio tangible, sino más bien de una secuencia de ceros y unos dentro un disco duro. Y al igual que al almacenar algo, interesa tenerlo bien organizado para optimizar el uso del espacio disponible lo máximo posible.
Aquí es donde entra la compresión de imágenes: en esa organización y optimización del espacio digital para tratar de no ocupar más capacidad de la necesaria. En este contexto, surgen distintos algoritmos que tienen la finalidad de conseguir reducir el tamaño de los archivos digitales tratando de mantener la calidad lo más cercana a la original posible.
Desde los años 90 el algoritmo omnipresente ha sido JPEG. La gran mayoría de imágenes que se generan cada día se guardan en este formato. Pero a pesar de estar tan extendido, tiene mucho margen de mejora aprovechando los avances en computación y el desarrollo de nuevos estándares y algoritmos.
JPEG XL es un novedoso estándar que trata de aprovechar muchas de las características de su predecesor, mientras añade muchas más, para que la compresión de imágenes a día de hoy sea más moderna, adecuada y eficiente.
Este es el objetivo de este trabajo. Tratar de implementar un algoritmo que se base en los principios de JPEG XL para comprimir imágenes. Se estudiará su eficiencia, calidad y resultados y se expondrán en este documento.
Para lograr este objetivo, se explora sobre las imágenes digitales en el contexto actual, sobre la luz, el color y las diferentes representaciones que se le dan al color en dispositivos digitales. Se aprenderá sobre los principios de JPEG y se conocerá su funcionamiento para poder conocer JPEG XL. Y por fin, se podrá implementar y evaluar el algoritmo que buscamos.
A digital image is a two-dimensional representation of an image in a matrix, which is stored and displayed by a computer or digital device. Like any stored object, it takes up space. However, here we are dealing with ”digital space”, where information is encoded as sequences of zeros and ones and stored on a hard drive. Just as we would arrange physical items for an efficient use of space, it is crucial to keep digital data well-organized in order to optimize available storage. This is where image compression comes into play: sorting and optimizing digital space to avoid using more capacity than necessary. In this context, various algorithms emerge with the aim of reducing the size of digital files while maintaining quality as closely as possible to the original. Since the 1990s, JPEG has been the go-to algorithm, with most daily generated images being saved in this format. Despite its widespread use, there is still plenty of room for improvement, especially with the advancements in computing and the development of new standards and algorithms. JPEG XL is a cutting-edge standard designed to build upon the features of its predecessor while introducing new ones, aiming to make image compression more modern, suitable, and efficient. That is the target of this study: to develop an algorithm based on the principles of JPEG XL for image compression. Its efficiency and quality will be evaluated, and results will be presented in this document. To accomplish this goal, we will thoroughly examine the current landscape of digital images, light, color, and the various color representations in digital devices. We will examine the principles of JPEG and its operation to gain a better understanding of JPEG XL. Then, we can finally implement these principles and evaluate the desired algorithm
A digital image is a two-dimensional representation of an image in a matrix, which is stored and displayed by a computer or digital device. Like any stored object, it takes up space. However, here we are dealing with ”digital space”, where information is encoded as sequences of zeros and ones and stored on a hard drive. Just as we would arrange physical items for an efficient use of space, it is crucial to keep digital data well-organized in order to optimize available storage. This is where image compression comes into play: sorting and optimizing digital space to avoid using more capacity than necessary. In this context, various algorithms emerge with the aim of reducing the size of digital files while maintaining quality as closely as possible to the original. Since the 1990s, JPEG has been the go-to algorithm, with most daily generated images being saved in this format. Despite its widespread use, there is still plenty of room for improvement, especially with the advancements in computing and the development of new standards and algorithms. JPEG XL is a cutting-edge standard designed to build upon the features of its predecessor while introducing new ones, aiming to make image compression more modern, suitable, and efficient. That is the target of this study: to develop an algorithm based on the principles of JPEG XL for image compression. Its efficiency and quality will be evaluated, and results will be presented in this document. To accomplish this goal, we will thoroughly examine the current landscape of digital images, light, color, and the various color representations in digital devices. We will examine the principles of JPEG and its operation to gain a better understanding of JPEG XL. Then, we can finally implement these principles and evaluate the desired algorithm
Description
Trabajo de Fin de Doble Grado en Ingeniería Informática y Matemáticas, Facultad de Informática UCM, Departamento de Arquitectura de Computadores y Automática, Curso 2023/2024