US 12,223,692 B2
Method and system for image-based reservoir property estimation using machine learning
Chicheng Xu, Houston, TX (US); Weichang Li, Katy, TX (US); and Yaser Alzayer, Dhahran (SA)
Assigned to SAUDI ARABIAN OIL COMPANY, Dhahran (SA)
Filed by ARAMCO SERVICES COMPANY, Houston, TX (US)
Filed on Jan. 20, 2022, as Appl. No. 17/580,079.
Claims priority of provisional application 63/140,491, filed on Jan. 22, 2021.
Prior Publication US 2022/0237891 A1, Jul. 28, 2022
Int. Cl. G06V 10/764 (2022.01); E21B 47/002 (2012.01); G01V 3/02 (2006.01); G01V 3/14 (2006.01); G06N 3/08 (2023.01); G06V 10/50 (2022.01); G06V 10/56 (2022.01); G06V 10/762 (2022.01); G06V 10/771 (2022.01); G06V 10/82 (2022.01)
CPC G06V 10/764 (2022.01) [E21B 47/002 (2020.05); G01V 3/02 (2013.01); G01V 3/14 (2013.01); G06N 3/08 (2013.01); G06V 10/507 (2022.01); G06V 10/56 (2022.01); G06V 10/763 (2022.01); G06V 10/771 (2022.01); G06V 10/82 (2022.01); E21B 2200/22 (2020.05)] 14 Claims
OG exemplary drawing
 
1. A method comprising:
obtaining, by a computer processor, core image data regarding a geological region of interest;
obtaining, by the computer processor, well log data regarding the geological region of interest from one or more wells;
determining, by the computer processor, a sliding window that corresponds to a predetermined window size;
determining, by the computer processor, a plurality of quantitative image attributes using the core image data, the well log data, and the sliding window,
wherein the plurality of quantitative image attributes are selected from a group consisting of color values, hue values, saturation values, and entropy values, and
wherein the plurality of quantitative image attributes are determined in a continuous manner by moving the sliding window along the core image data; and
generating, by the computer processor, predicted rock data for the geological region of interest using the plurality of quantitative image attributes, a machine-learning algorithm, and a machine-learning model,
wherein the machine-learning algorithm comprises a K-means clustering algorithm that outputs the machine-learning model, and
wherein the machine-learning model comprises a plurality of clusters that are organized according to one or more visual rock types (VRTs) that are indexed by depth.