New computational techniques for optimal signal/image processing
(A Torokhti)
Most hi-tech electronic devices must process signals. E.g., a mobile phone
must encode, transmit, receive and decode voice signals. An image can be
transformed (by a digital processor) in a collection of digital signals.
Therefore, an image is treated as a signal. This project will use
specialised mathematical theories applied in novel ways to advance the
theoretical foundations of signal/image processing and develop better
processing algorithms for practical applications. Companies with access to
better signal processing algorithms have an edge over their competitors, and
consumers benefit too from better and more advanced products.
This is an ambitious project that requires highly innovative ideas and techniques, and therefore is capable of attracting future external funding. The research team members are among the most well known world researchers in the field. Although the project initiates a new research direction, it will extend existing results (including those of the team) published in relevant journals with the highest impact factor. This evidences the project's high quality.
Funding
2008: University of South Australia, Division of Information Technology, Engineering and Environment Research Grant. Project title: "New computational techniques for optimal signal/image processing", $10,000 (Chief Investigator: A/Prof. A. Torokhti).
Publications
A. Torokhti and P. Howlett, Computational Methods for Modelling of Nonlinear Systems, Elsevier, 397 p., 2007.
A.P. Torokhti, P. G. Howlett and C. Pearce, New Perspectives on Optimal Transforms for Random Vectors, Optimization: Theory and Applications, Springer, 2009 (accepted).
S. Friedland and A. P. Torokhti, Generalized rank-constrained matrix approximations, SIAM J. Matrix Anal. Appl., 29, issue 2, pp. 656-659, 2007.
A. Torokhti and S. Friedland, Towards theory of generic Principal Component Analysis, J. Multivariate Analysis, (accepted).
A. Torokhti and P. Howlett, Filtering and Compression for Infinite Sets of Stochastic Signals, Signal Processing, (accepted).
P. G. Howlett, A. Torokhti and C.E.M. Pearce, Optimal Multilinear Estimation of a Random Vector under Constraints of Causality and Limited Memory, Computational Statistics and Data Analysis, 52 , Issue 2, pp. 869-878, 2007.
A. Torokhti and S. Miklavcic, Compression and Filtering of Random Signals under Constraint of Variable Memory, Proc. of World Academy of Sci., Eng. & Tech. Vol. 30, July 2008, ISSN 1307-6884, pp. 768-773.
A. Torokhti and P. Howlett, Optimal Data Compression and Filtering: the Case of Infinite Signal Sets, Proc. of World Academy of Sci., Eng. & Tech. Vol. 30, July 2008, ISSN 1307-6884, pp. 774-783.
A. Torokhti, S. Friedland and P. Howlett, Towards generic theory of data compression, CD-rom: 2007 IEEE Int. Symp. on Inform. Theory. June 24-29 2007, Nice, France. Symposium Proceedings. ISBN: 1-4244-1429-6, pp. 291-295, 2007.
A. Torokhti, P. Howlett and C. Pearce, Information estimation from partially missed data, CD-rom: 2007 IEEE Int. Symp. on Inform. Theory. June 24-29 2007, Nice, France. Symposium Proceedings. ISBN: 1-4244-1429-6, pp. 2011-2015, 2007.