Research On Extraction And Application Of Features Of Electro-imaging Log Image

Carbonatite reservoirs have various types of reservoir space and are highly heterogeneous,many experts and scholars have carried out in-depth research on reef reservoir logging evaluation.But most are simply stratigraphic and qualitative interpretation based on conventional well logs.Relying on conventional logging is difficult to distinguish between lithology and sedimentary facies,electro-imaging logging data can clearly reflect the structural components and sedimentary structures of carbonate reservoirs,it provides a reliable basis for distinguishing lithology and sedimentary facies.Electrical imaging logging data can visually distinguish different geological features.However,at present,most of the experts and scholars have stayed on the artificially processing of the application of the electrical imaging logging data.In this paper,the use of electrical imaging logging data to extract relatively features to divide layers,and studying the correlation between the geological characteristics of the non reservoir and the sedimentary model.This paper focuses on the texture feature extraction and application of electrical imaging logging images,mainly extracted Graylevel co-occurrence matrix texture feature,Tamura texture feature,Laws texture feature,and resistivity spectrum feature of electrical imaging logging images.The features extracted from the above methods are compared and analyzed,combined with electrical imaging logging,static and dynamic map,apparently most of the characteristic curves reflect the formation information,then use the SOM algorithm to automatically stratigraphy.Then the paper analyzes the response of non-reservoir geological characteristics on electrical imaging logging and conventional logging data,puts forward 6 sedimentary log interpretation models which are related to the related geological characteristics.Then extracts the shape features of pictures,the minimum Euclidean distance algorithm and extracted image shape features are used to calculate and match the corresponding sedimentary well interpretation models.The algorithms of this paper are written by C language,the above module is hooked up to the LEAD logging comprehensive interpretation software,and having processed actual logging data,and got excellent application effect.