Method and device for automatically extracting image features of electric imaging logging

Abstract

The invention provides a method and a device for automatically extracting image features of electric imaging logging, wherein, the method comprises the following steps: acquiring historical data of electric imaging logging; Pre-processing the historical data of electric imaging logging to generate the electric imaging logging image covered by the whole wellbore; The typical geological features areidentified and marked in the electric imaging logging images covered by the whole borehole, and the processed images are obtained, and the processed images are determined as training samples according to the geological feature types. Constructing a depth learning model, the depth learning model comprising an input layer, a plurality of hidden layers and an output layer; Using training samples totrain the depth learning model; The trained depth learning model is used to identify the geological feature types of the electric imaging logging images of the well sections to be identified, and themorphological optimization processing is carried out on the recognition results to obtain the feature optimization recognition results. This scheme can automatically, quickly and accurately identify the typical geological features in electrical imaging logging images.

Description

Technical field [0001] The invention relates to the technical field of electrical imaging logging data processing and interpretation, in particular to a method and device for automatically extracting electrical imaging logging image features. Background technique [0002] Electrical imaging logging is currently the most widely used imaging logging method. The electrical imaging logging images can visually display shale, bioclastics, calcite crystals, pores, fractures, bedding, faults, folds and other stratum deposition and structural features. Provide an important basis for comprehensive reservoir analysis. Traditional electrical imaging logging image interpretation is mainly done by experts directly qualitatively distinguishing various features based on experience. It relies heavily on expert experience and interpretation is highly subjective, which is difficult to meet the urgent needs of oilfield exploration and production. How to accurately identify, classify, and quantitati...

AI Technical Summary

Problems solved by technology

Although this type of method has also achieved certain results in practical applications, there are still the following problems: 1) The effect of feature extraction is controlled by the quality of image segmentation. Accurate separation of image features leads to missing or misjudgment of feature extraction; 2) The classification method based on feature parameter statistics is affected by the selected parameter indicators, and the recognition accuracy is limited, and it is usually related to the geological conditions of a specific area, and its universality is poor; 3) Now Existing technologies usually only identify and extract specific geological feature targets, such as fractures and caves, which cannot meet the needs of differentiation and diversity in geological research and reservoir evaluation

Images