Lithology identification method under microscope based on convolutional neural network

Abstract

The invention relates to classification and identification of rock lithology, in particular to a microscopic lithology identification method based on a convolutional neural network, which comprises the following steps: acquiring a plurality of rock slice microscopic images to be identified as a sample set; inputting the sample set into a trained convolutional neural network, and outputting classification information of the sample set, wherein a rock slice microscopic image set is obtained by adopting orthogonal polarized light or single polarized light or both of the orthogonal polarized lightand the single polarized light, and the rock slice microscopic image set forms the sample set. According to the method, a computer is used for automatically obtaining feature description of the images through learning and automatically classifying the images, the labor cost and the learning cost are remarkably reduced, the lithology identification speed is greatly increased, and the method has the advantages that the oil and gas exploration and development benefits are improved efficiently and conveniently, and the development cost is reduced.

Description

technical field [0001] The invention relates to the classification and identification of rock lithology, in particular to a method for lithology identification under a microscope based on a convolutional neural network. Background technique [0002] The identification of lithology in geological work is firstly identified with the naked eye and a magnifying glass. However, the naked eye identification results are often not accurate enough, so it is necessary to bring the rock sample back to the laboratory to grind it into a rock thin section for observation and description under a polarizing microscope. The lithology can be comprehensively judged by determining its mineral composition, relative content, structure composition and other optical forms under a microscope. For any geologist, the technology of identification under the microscope is the basis of all work, but the manual identification of rock thin sections under the microscope requires a large amount of mineral kno...

AI Technical Summary

Problems solved by technology

For any geologist, the technology of microscopic identification is the basis of all work, but manual identification of rock thin sections under the microscope requires a large amount of mineral knowledge for the appraiser in the early stage, and the workload of microscopic identification is large, and the cost of manpower and material resources is high. Coupled with the gap in everyone's cognition, there is often a certain gap in the appraisal results

Images