Visual Recognition of Drill Cuttings Lithologies Using Convolutional NeuralNetworks to Aid Reservoir Characterisation

摘要

There are a vast number of reservoirs with drill cuttings and core images that have classification problemsassociated with them.This could be due to the images not being classified in the first place,or the imagesmay be available but the interpretation reports could be missing.Another problem is that images fromdifferent wells could be interpreted by different wellsite geologists/sedimentologists and hence result in aninconsistent classification scheme.Finally,there could also be the problem of some images being incorrectlyclassified.Ergo it would be desirable to have an unbiased objective system that could overcome all ofthese issues.Step in convolutional neural networks.Advances during this decade in using convolutionalneural networks for visual recognition of discriminately different objects means that now object recognitioncan be achieved to a significant extent.Once the network is trained on a representative set of lithologicalclasses,then such a system just needs to be fed the raw drill cuttings or core images that it has not seenbefore and it will automatically assign a lithological class to each image and an associated probabilityof the image belonging to that class.In so doing,images below a certain probability threshold can beautomatically flagged for further human investigation.The benefit of such a system would be to improvereservoir understanding by having all available images classified in a consistent manner hence keeping thecharacterization consistent as well.It would further help to reduce the time taken to get human expertise tocomplete the task,as well as the associated cost.