AUTOMATED METHODOLOGY FOR DETECTING BORDER IN CT-SCAN IMAGES OF NON-CIRCULAR ROCK SAMPLES

Abstract

X-ray computed tomography, CT-scan, is a very important tool in special core analysis, because it allows the visualization of the interior of cores, providing information such as porosity, saturations, and changes on these parameters with time. CT-scan images are usually square, and contain information not only the core, but also parts of the core-holder, the water used to pressurize the system, and the rubbers that hold the cores in place. Software available to process these images are usually very expensive, or do not allow reasonable automation of the process of treating images. This paper proposes a simplified methodology to isolate the core in a CT image, allowing the processing of multiple images without the interference of the user. This methodology can be implemented easily in any computer language, it saves operation time, and reduces the risk of human error.