PDF version of this document

Overview

HROUGHOUT the past year, many results have been obtained and better understanding established. Some of these are more relevant to the preliminary research aims (see Form 2) than others. Although experiments were applied to 1-D data, principles that extend to a higher-dimensional data were learned and should form the theoretical grounds for future work concerning the model-based objective function. It has been agreed that shortly into the second year of this research, once sufficient understanding of the problem and confidence are gained, 2-D (and possibly later on, 3-D) application of the method will be investigated^7.1.

Indication of prescribed tasks and some intermediate submissions of progress up-to-date can be found in the forms. They are available online^7.2 as Portable Document Format (PDF) and Word files at:

• http://www.danielsorogon.com/Webmaster/Research/Progress/forms.htm

All weekly progress reports can be found under:

• http://www.danielsorogon.com/Webmaster/Research/Progress/

There are clear advantages to the retaining of the records above. These will later allow chronological dissection of progress made, as well as the problems encountered on a day-to-day or week-to-week basis. They are yet not adhesive enough and this chapter attempts to provide a short summary of the large bulk of experiments by just listing some of the more important ones and explaining how they serve the raison d'être of the project. All experiments are documented to their finest detail at:

• http://www2.cs.man.ac.uk/ schestr0/Experiments/

They were generated using AART which is a newly-constructed tool that is shown in Figure cap:Autonomous-Appearance-based-Registration and explained in Section sec:Environment.

All results were obtained under MATLAB [WWW-1] which is the working environment on which AART operates^7.3.