The paper describes the fracture recognition methods and algorithms for automated readout of experimental data (openness, intensity, extent, and surface density of fractures) from petrographic thin section images and photographs of core. Previously, the experimental data, such as length, openness, and surface density of fractures, was read off manually from core cuts and required much time.
Some of the most important properties of core can be determined by analyzing digital images of petrographic thin sections and core pictures. At the same time, assessment of the properties of fractured-vuggy reservoirs in this case is not an easy task and the creation of a tool for automated estimation of core fracture parameters is a topical issue today.
This issue is addressed through a modern automated approach to the processing of experimental data obtained from studies of luminophore-saturated core samples. The advancement in computer technologies has allowed to develop digital methods for obtaining and analyzing optical images of core samples that provide processing full-size information with a high degree of accuracy.
The method of finding these indicators in computer processing is to identify the boundaries of fractures on the surface of core thin sections. One of effective algorithms of the SUSAN boundary detector (Smallest Univalue Segment Assimilating Nucleus) is used for this. To determine the geometric length of a fracture, image thinning operation based on Zhang – Suen algorithm is applied. An original algorithm for measuring the width and extent of fractures with a specified error has also been developed.
The algorithm for recognizing and determining the size of fractures from thin section images can be divided into four stages. On the first stage the area of the studied object in the picture is determined. The objects in the photos that can determine fracture parameters - photos of thin sections, photos of full-sized core - have irregular geometric dimensions. Therefore, when studying the fracturing in the images provide a tool for correct and accurate estimation of the object area in the photograph. The second stage is an estimation of the length of all fractures in the photo. To estimate the fracturing parameters, the lengths of all the fractures in the picture, taking into account their tortuosity, should be included. On the third stage the width of fractures is estimated. To estimate the fracture capacity (fracture porosity), it is necessary to evaluate the openness of each fracture and find the average of this parameter. The fourth stage is fracturing parameters estimate. Using the values obtained fracture we can estimate porosity and rock permeability parameters.
The advantages of software processing of experimental data include the computation rate, the accuracy of determining the geometric characteristics close to manual estimation, and the possibility of further increase of the program functions. Additional value includes the possibility of automating the preliminary analysis of core material and readout of experimental data from petrographic thin section images and core photographs, as well as rapid evaluation of fracture geometry in reservoirs.
References
1. Bagrintseva K.I., Sautkin R.S., Shershukov G.I., Metodika programmnoy obrabotki eksperimental'nykh dannykh posle nasyshcheniya karbonatnykh porod lyuminoforom (Technique of program processing of experimental data after carbonate rocks saturation with a phosphor), Proceedings of III International Conference of Young Scientists and Specialists “Aktual'nye problemy neftegazovoy geologii KhKhI veka” (Actual problems of oil and gas geology of the XXI century), Part 4, St. Petersburg, Publ. of VNIGRI, 2013, pp. 4–7.
2. Klyuev A.V., Aristov G.V., Opredelenie parametrov mikrostruktury metallov metodami komp'yuternogo zreniya (Determination of the parameters of the microstructure of metals by computer vision methods), Proceedings of XII All-Russian school-conference of young scientists “Upravlenie bol'shimi sistemami” (Managing large systems), Volgograd, 2015, pp. 701–714.
3. Taylakov O.V., Makeev M.P., Algorithmic support of the analysis of optical images of anschlift-ore and its application for evaluating structural changes in coals (In Russ.), Gornyy informatsionno-analiticheskiy byulleten' (nauchno-tekhnicheskiy zhurnal), 2008, V. S13, pp. 189–197.
4. Gmid L.P., Metodicheskoe rukovodstvo po litologo-petrograficheskomu i petrokhimicheskomu izucheniyu osadochnykh porod-kollektorov (Methodological guidelines for lithologic-petrographic and petrochemical studies of sedimentary reservoir rocks), St. Petersburg, Publ. of VNIGRI, 2009, 160 p.
5. Ivanov D.V. et al., Algoritmicheskie osnovy rastrovoy grafiki (Algorithmic fundamentals of raster graphics), URL: http://www.intuit.ru/goto/course/rastrgraph/
6. Smith S.M., Brady J.M., SUSAN – a new approach to Low Level Image Processing, DRA Technical Report TR95SMMS1b, 1995, 57 p.
7. Molchanova V.S., Eight-connected asymmetric skeletonization algorithm for binary images (In Ukr.), Vіsnik SumDU. Serіya “Tekhnіchnі nauki”, 2013, no. 2, pp. 43–50.
