Wavelet-based image compression for mobile applications.

The transmission of digital colour images is rapidly becoming popular on mobile telephones, Personal Digital Assistant (PDA) technology and other wireless based image services. However, transmitting digital colour images via mobile devices is badly affected by low air bandwidth. Advances in communications Channels (example 3G communication network) go some way to addressing this problem but the rapid increase in traffic and demand for ever better quality images, means that effective data compression techniques are essential for transmitting and storing digital images. The main objective of this thesis is to offer a novel image compression technique that can help to overcome the bandwidth problem. This thesis has investigated and implemented three different wavelet-based compression schemes with a focus on a suitable compression method for mobile applications.
The first described algorithm is a dual wavelet compression algorithm, which is a modified conventional wavelet compression method. The algorithm uses different wavelet filters to decompose the luminance and chrominance components separately. In addition, different levels of decomposition can also be applied to each component separately. The second algorithm is segmented wavelet-based, which segments an image into its smooth and nonsmooth
parts. Different wavelet filters are then applied to the segmented parts of the image. Finally, the third algorithm is the hybrid wavelet-based compression System (HWCS), where
the subject of interest is cropped and is then compressed using a wavelet-based method. The details of the background are reduced by averaging it and sending the background separately from the compressed subject of interest. The final image is reconstructed by replacing the averaged background image pixels with the compressed cropped image.
For each algorithm the experimental results presented in this thesis clearly demonstrated that encoder output can be effectively reduced while maintaining an acceptable image visual quality particularly when compared to a conventional wavelet-based compression scheme.

A thesis submitted to Middlesex University in partial fulfîlment of the requirements for the degree of Doctor of Philosophy.