BEAMS

Principal Investigator: Dr  M Van Der Baan

Sponsor: Shell International

Value: £67000.00

Dates: 1st March 2007 to 29th February 2008

Summary

The project aims to introduce a step-change in the multiple subtraction problem. Desired primaries are to be retrieved from the recorded data given an imperfect prediction of the multiples. The prediction suffers from amplitude and timing inconsistencies and a residual wavelet may be present. The objective is to subtract the multiples from the data adaptively such that the desired primaries are recovered without further a prior information - hence the term blind in the project title. Desired primaries will be isolated from the multiples by means of an overcomplete independent component analysis (ICA), possibly combined with a blind deconvolution approach to correct for the residual wavelet.