Date of Award:
Master of Science (MS)
Stephen J. Allan
Daniel W. Watson
This paper examines the application of genetic programming to image compression while working in the frequency domain. Several methods utilized by JPEG encoding are applied to the image before utilizing a genetic programming system. Specifically, the discrete cosine transform (DCT) is applied to the original image, followed by the zig-zag scanning of DCT coefficients. The genetic programming system is finally applied to the one-dimensional array resulting from the zig-zag scan. The research takes an existing genetic programming system developed for the spatial domain and develops DCT domain functionality. The results from the DCT domain-based genetic programming system are compared with those from the spatial domain-based system, and show improvements to the image quality with a reduction up to half of the evolved image's average error. The results show that working in the frequency domain has advantages over the spatial domain. Several methods to exploit these advantages are proposed and evaluated.
Banham, Benjamin E., "An Evolutionary Approach to Image Compression in the Discrete Cosine Transform Domain" (2008). All Graduate Theses and Dissertations. 5.
Copyright for this work is retained by the student. If you have any questions regarding the inclusion of this work in the Digital Commons, please email us at .