Image enhancement using coherence processing with applications to seismograms

This research discusses the application of one particular coherence measure in two-dimensional image processing of seismograms. The coherence measure provides a quantitative method of estimating the similarity of the data among the channels in multichannel data. Estimates of directional coherence, expressed in signal-to-noise ratio, are applied as weights in a processing technique by which images are enhanced. The estimator and enhancement algorithms are presented and the estimator properties are discussed. Both synthetic seismograms and field data are processed using the event detection, SNR estimation and enhancement/extraction capabilities of the processing technique. Though the technique has been succesfully applied to both vertical [1] and horizontal profile data, only the horizontal profiling data are treated here.