University Librarian Centre | University of Cagliari
UniCA Eprints

JPEG XR scalable coding for remote image browsing applications

Saba, Bernardetta (2012) JPEG XR scalable coding for remote image browsing applications. [Doctoral Thesis]



The growing popularity of the Internet has opened the road to multimedia and interactivity, emphasizing the importance of visual communication. In this context, digital images have taken a lead role and have an increasing number of applications. Consider, for example, the spread that digital cameras and mobile devices such as mobile phones have become in recent years. Thus, it arises the need for a flexible system that can handle images from different sources and are able to adapt to a different view. The importance of this issue lies in the application scenario: today there are datastores with a large number of images saved in JPEG format and systems for rendering digital images are various and with very different characteristics with each other. The ISO/IEC committee has recently issued a new format, called JPEG-XR, created explicitly for the modern digital cameras. The new coding algorithm JPEG-XR, can overcome various limitations of the first JPEG algorithm and provides viable alternatives to the JPEG2000 algorithm. This research has primarily focused on issues concerning the scalability of the new format of digital images.Additional scalability levels are fundamental for image browsing applications, because enable the system to ensure a correct and efficient functioning even when there is a sharp increase in the number of resources and users.Scalability is mostly required when dealing with large image database on the Web in order to reduce the transferred data, especially when it comes to large images. The interactive browsing also requires the ability to access to arbitrary parts of the image. The starting point is the use of a client-server architecture, in which the server stores a database of JPEG XR images and analyzes requests from a client. Client and server communicate via HTTP and use an exchange protocol. In order to minimize the transferred information, the JPEG XR coded file format should make use of the frequency mode order and partitioning of images into optimized tiles. The main goal is transmitting only some subset of the available sub-band coefficients. This is necessary to allow access an interactive access to portion of images, that are downloaded and displayed, minimizing the amount of data transferred and maintaining an acceptable image quality.The proposed architecture has of course prompted a study of errors in transmission on unreliable channel, such as the wireless one, and the definition of possible optimizations/variants of the codec in order to overcome its own limitations. Image data compressed with JPEG XR when transmitted over error-prone channels is severely distorted. In fact, due to the adaptive coding strategies used by the codec, even a single bit error causes a mismatch in the alignment of the reading position from the bit-stream, leading to completely different images at the decoder side. An extension to the JPEG XR algorithm is proposed, consisting in an error recovery process enabling the decoder to realign itself to the right bit-stream position and to correctly decode the most part of the image. Several experiments have been performed using different encoder parameter and different error probabilities while image distortion is measured by PSNR objective metric. The simplicity of the proposed algorithm adds very little computational overhead and seems very promising as confirmed by objective image quality results in experimental tests.

Item Type:Doctoral Thesis
Date:20 March 2012
Tutor:Giusto, Daniele
PhD classes:Ciclo 23 > Ingegneria elettronica e informatica
Coordinator:Giua, Alessandro
Institution:Universita' degli Studi di Cagliari
Divisions:Dipartimenti (fino a dicembre 2011) > Dipartimento di Ingegneria elettrica ed elettronica
Subjects:Area 09 - Ingegneria industriale e dell'informazione > ING-INF/03 Telecomunicazioni
Uncontrolled Keywords:JPEG XR, codifica di immagini, canali con errore, image coding, error channel
ID Code:770
Deposited On:05 Apr 2012 12:22

Repository Staff Only: item control page