Document Type

Theses, Masters

Master Thesis

Master thesis

Rights

Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence

Disciplines

Electrical and electronic engineering

Publication Details

A Thesis Presented For the Award of Masters by Research, Technological University Dublin, January 2021.

Abstract

Remote working is becoming increasingly more prevalent in recent times. A large part of remote working involves sharing computer screens between servers and clients. The image content that is presented when sharing computer screens consists of both natural camera captured image data as well as computer generated graphics and text. The attributes of natural camera captured image data differ greatly to the attributes of computer generated image data. An image containing a mixture of both natural camera captured image and computer generated image data is known as a compound image. The research presented in this thesis focuses on the challenge of constructing a compound compression strategy to apply the ‘best fit’ compression algorithm for the mixed content found in a compound image. The research also involves analysis and classification of the types of data a given compound image may contain. While researching optimal types of compression, consideration is given to the computational overhead of a given algorithm because the research is being developed for real time systems such as cloud computing services, where latency has a detrimental impact on end user experience. The previous and current state of the art videos codec’s have been researched along many of the most current publishing’s from academia, to design and implement a novel approach to a low complexity compound compression algorithm that will be suitable for real time transmission. The compound compression algorithm will utilise a mixture of lossless and lossy compression algorithms with parameters that can be used to control the performance of the algorithm. An objective image quality assessment is needed to determine whether the proposed algorithm can produce an acceptable quality image after processing. Both traditional metrics such as Peak Signal to Noise Ratio will be used along with a new more modern approach specifically designed for compound images which is known as Structural Similarity Index will be used to define the quality of the decompressed Image. In finishing, the compression strategy will be tested on a set of generated compound images. Using open source software, the same images will be compressed with the previous and current state of the art video codec’s to compare the three main metrics, compression ratio, computational complexity and objective image quality.

DOI

https://doi.org/10.21427/syps-p748


Share

COinS
 

Document Type

Master thesis