Introduction

 

 

 

 

 

 

 

 

 

 

 

There are many areas of scientific data processing in which the data is non-uniformly distributed. Beyond resampling the data to a regular grid there are currently few techniques for processing such data. This project aims to develop a general framework for processing irregular mesh data that copes with problems such as points having different numbers of neighbours, at varying distances, and non-uniform sampling. In addition, uncertainties in the data should be propagated through the processing pipeline. Methods developed by the project will enable scattered data to be analysed and edited directly, without requiring resampling on a regular grid, with the potential for more efficient and accurate processing of scattered data in many application areas. Using the above framework the project will develop algorithms for low-level operations (namely noise filtering and binary operations) that are an essential pre-processing step to higher-level mesh processing. As well as algorithms being developed for specific low-level operations, the project has wider applicability since the general approaches developed will potentially be of use in transferring other image processing operations to mesh processing.

Funded by