Date of Award

1-2015

Document Type

Master Thesis

Degree Name

Master of Science

Department

Computer Sciences

First Advisor

Paul Walsh

CIT Disciplines

1.2 COMPUTER AND INFORMATION SCIENCE

Abstract

This thesis presents novel systems for the automatic and semi-automatic design of digital artefacts. Currently, users wanting to create digital models, such as three-dimensional (3D) digital landscapes and website colour schemes, need to possess significant expertise, as the tools involved demand a high level of knowledge and skill. By developing an intuitive algorithmic process, founded on evolutionary computation (EC), this research enables non-specialist human designers to create digital assets more efficiently. This is achieved by replacing design activities that require significant manual input with algorithmic functions, thereby greatly improving the efficiency and accessibility of the practices involved.

This research places an initial focus on the generation of 3D landscapes, but the latter aspect concentrates on the identification of text and background colour combinations more amenable to the reading process, particularly for readers with vision impairments. Choosing an ideal combination of colours requires knowledge of the cognitive and psychological procedures involved. Designers need to be aware of colour contrast ratios, brightness, and variations, which would require a series of aesthetic measurements if they are to be manually tested. In an effort to provide a colour design facility, this research offers algorithms that can generate colour schemes, based on the aforementioned principles, which can be used to derive an optimum scheme for a website.

This research demonstrates a novel interactive genetic algorithm (IGA), coupled with the use of computational aesthetics, suitable for use in the evolution of terrain generation and digital landscape design. It also provides a tool for automatically creating EC-driven colour palettes for web design via evolutionary searches.

Experimental trials use the EC framework developed from this research using both IGA technique and the computational aesthetic measures. Results indicate that the end-users can build any target digital landscape design with less inputs and more comfort, and if required can also automate the whole process to evolve aesthetically pleasing landscape designs. The results obtained for designing colour schemes for website design have proven that end-users can quickly develop a colour scheme, without the need for fine-tuning of colour combinations. Results can compete in quality the colour schemes that are designed by the professional website developers.

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Access Level

info:eu-repo/semantics/openAccess

Download

Share

COinS