Image Processing and Machine Learning Approaches for Petrographic Thin Section Analysis

Abstract

The article presents the methodology of petrographic thin section analysis, combining the algorithms of image processing and statistical learning. The methodology includes the structural description of thin sections and rock classification based on images obtained from polarized optical microscope. To evaluate the properties of structural objects in thin section (grain, cement, voids, cleavage), first they are segmented by watershed method with advanced noise reduction, preserving the boundaries of grains. Analysis of segmentation for test thin sections showed a fairly accurate contouring of mineral grains which makes possible automatically carry out the calculation of their key features (size, perimeter, contour features, elongation, orientation, etc.). The paper presents an example of particle size analysis – definition of grains size class. The roundness and rugosity coefficients of grains are estimated also. Statistical analysis of templates for manual determination of roundness and rugosity coefficients revealed drawback of examined templates in terms statistical accuracy (high dispersion of coefficient for all grain within one template, outliers presence). In the frame of classification problem the feature importance analysis and clustering of non-correctly segmented grains are handled. The classifier for rock type definition (sandstone, limestone, dolomite) is trained with decision tree method, while the classifier of mineral composition of sandstones (greywackes, arkose) is learnt with "random forest" method. Both classifiers are learnt in the feature space generated from segmented grains and their evaluated properties. As a result, we proved the possibility to conduct automatic quantitative and qualitative analysis of thin sections applying image processing and statistical learning methods.

Documenting archaeological thin sections in high‐resolution: A comparison of methods and discussion of applications

“…Multiple types of image analyses ought to benefit from this, such as basic raster manipulation and various types of quantitative image analyses (Goldberg & Macphail, ) to more sophisticated and novel approaches, such as machine learning (Budennyy et al, ; Ross, Fueten, & Yashkir, ). The power of digital micromorphology becomes particularly evident when thin section‐wide imagery is implemented into georeferenced investigative frameworks, where in combination with other archaeological datasets, they are capable of bridging the gap between microscale and macroscale investigations of archaeological contexts (Fisher et al, ; Haaland, Friesem, Miller, & Henshilwood, ; Karkanas et al, ).…”

Superpixel segmentations for thin sections: evaluation of methods to enable the generation of machine learning training data sets

“…1b), a classifier is trained on top of images where target mineral grains are manually identified ahead of time. The output is usually a segmentation map where each type of mineral is indicated by a unique color-mode (Budennyy et al, 2017;Thompson et al, 2001;Baykan and Yılmaz, 2010;Borges and de Aguiar, 2019;Ramil et al, 2018). Compared to image tagging, generating training sets for grain segmentation and identification is more time-consuming, as it requires a detailed mineral detection and identification process.…”

Evaluating the Effectiveness of an Artificial Intelligence Model for Classification of Basic Volcanic Rocks Based on Polarized Microscope Image

“…인공지능 영상처리기법은 공극이나 결정질 광 물들의 이미지상에서 나타나는 특성을 추출하여 광물 식 별 및 암석 분류 등에 활용한다 (Borazjani et al, 2016). 선행연구에서 개발된 모델은 화성암에서 광물 입자와 공 극을 구분하고, 각각의 광물 별로 편광현미경 이미지상 에서 보이는 특성을 추출하여 광물을 식별할 수 있다 (Izadi et al, 2013, Budennyy et al, 2017, Borges and de Aguiar, 2019…”