Report From Oslo Conference

www.open.ac.uk SCALABLE MULTIVIEW IMAGING

SCIA 2009 Report on Workshop & Tutorial BY: MOBOLAJI AKINOLA SUPERVISORS: PROF. LAURENCE DOOLEY DR. PATRICK WONG

©2009 by Mobolaji Akinola

Workshop On Fusion Techniques • Motivation: – Image Analysis Task hard to solve from single Image • Examples: –Traffic Surveillance, visible no. plates? –Medical Imaging , NMR (Nuclear Magnetic Resonance) gives certain details while PET (Positron Emission Tomography) can give others – Low Resolution Images can be combined to get better resolution?? – Combination of different modes and modalities exploiting where each is better?? • Conclusion – [1.] More images [2.] Combination techniques ©2009 by Mobolaji Akinola

Workshop On Fusion Techniques • Fusion Categories: – Mutiview Fusion, complementary info views – Mutitemporal Fusion, diff views, same scene, diff times » Goal – intermediate stage synthesis

– Multimodal Fusion, Emphasize band specific information » Methods- pixel averaging, object level fusion, transform domains

– Multisetting Fusion, multifocus, different exposure etc – Mutichannel deconvolution, blurr due to diff. convolutions – superesolution

©2009 by Mobolaji Akinola

Multiview fusion Metrodome

B. Zitová and J. Flusser, Image registration methods: A survey, Image and Vision Computing, 21(11): 977-1000, 2003

Multitemporal fusion digital subtraction angiography

Courtesy of Y. Bentoutou et al.

Multimodal fusion medical imaging – pixel averaging

NMR

SPECT

Workshop On Image Quality Measurements • • • •

Present Measures inadequate PSNR Vs Subjective Measure Modeling Human Visual System input of statistical experiments

• Conclusion: – Other metrics at infancy – Promising ongoing experiments ©2009 by Mobolaji Akinola

THANK YOU

©2009 by Mobolaji Akinola