Evaluation of segmentation procedures using X-ray computed microtomography images of coquinas from Morro do Chaves Formation - NE Brazil
Fernanda O. Hoerlle1; William G. A. L. Silva1; Edmilson H. Rios1; Thaís M. G. Silveira1; Paulo Couto1; José L. D. Alves1; Maira C. O. Lima1; Leonardo Borghi2; Patrick W. M. Corbett3
1 Programa de Engenharia Civil - COPPE/Universidade Federal do Rio de Janeiro; 2 Laboratório de Geologia Sedimentar (Lagesed), Instituto de Geociências, Universidade Federal do Rio; 3 Heriot-Watt University
doi:10.20906/CPS/CILAMCE2017-0573
Resumo
In recent years, the petroleum industry started to increase the use of digital image techniques, as X-ray computed microtomography (µCT), aiming the characterization of reservoir rocks. A frequent challenge is to determine the threshold value that binarizes the rock image into pore and matrix (solid phase). In this study, three coquinas samples from Morro do Chaves Formation, which are considered an analogue to the Pre-salt reservoirs, were analyzed aiming the quantification of their porosity through µCT images. This work compares manual and different automated segmentation methods using the µCT images with two different pixel sizes, 18µm and 10µm. The automated segmentation methods have well-defined computational algorithms that automatically select the threshold value and isolate the porous space and the matrix of the rock. The codes used were clustering-based, minimum error-based and local-based methods as well as an active contour without edges procedure. The porosity results of different segmentation techniques were compared with the one obtained with a helium porosimeter, which is the industry standard procedure to measure the porosity of rocks. Comparing all the data it was possible to define the best image resolution and segmentation technique for these coquina rocks.
Palavras-chave: Binarization ; Carbonate rocks; Digital Petrophysics; Porosity; Digital Image Techniques