Opportunities for data fusion in multi-modality tomography

Robert M. West, Richard A Williams

Research output: Chapter in Book/Report/Conference proceedingConference contribution

23 Citations (Scopus)

Abstract

It is now possible to fit several tomographic systems to an industrial process. By the use of different modalities to image the process, the process engineer can plan to take advantage of the strongest features provided by each unit. Detailed information from different modalities might be gained by inspection of separate tomograms. To maximise the information available from the combined tomographic system however data fusion is clearly the better option. A combined tomogram can be produced of superior quality to any of the separate tomograms: the sum can be better than the parts. Fusion might be achieved at different levels. Greater accuracy may be achieved by fusion at the measurement level, but at the expense of centralised processing. An example is given for a hydrocyclone where data from an electrical resistance tomograph and a reflective-mode ultrasound tomograph are fused. A discussion of issues relevant for fusion in process tomography is given.

Original languageEnglish
Title of host publication1st World Congress on Industrial Process Tomography
PublisherInternational Society for Industrial Process Tomography
Pages195-200
Number of pages6
ISBN (Print)9780853163176
Publication statusPublished - 1999
Event1st World Congress in Industrial Process Tomography - Manchester, United Kingdom
Duration: 14 Apr 199917 Apr 1999

Conference

Conference1st World Congress in Industrial Process Tomography
Country/TerritoryUnited Kingdom
CityManchester
Period14/04/9917/04/99

Keywords

  • Data fusion
  • Hydrocyclone

ASJC Scopus subject areas

  • Computational Mechanics
  • Control and Systems Engineering
  • Computer Science Applications
  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Opportunities for data fusion in multi-modality tomography'. Together they form a unique fingerprint.

Cite this