Accurate and automated detection of fractures in borehole image using YOLO-V8 model DOI
Yiming Yang, Qingchun Li, Hui Li

et al.

Petroleum Science and Technology, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 19

Published: Nov. 19, 2024

The digital panoramic drilling camera system has produced numerous high-precision borehole images essential for analyzing subsurface geological structures. However, manual segmentation and preprocessing of these are labor-intensive subjective. This paper introduces an innovative approach using the YOLOv8 model, trained on a substantial dataset fractures, to achieve fast, stable, automated recognition. model effectively learns extracts fracture features, accurately identifying defects even in complex contexts such as developed rock cleavage, schistosity, quartz vein bands, mixed alteration. Experimental results indicate that while is slightly slower inference speed than YOLOv5, it excels pixel accuracy, recall rate, mean Average Precision (mAP) segmentation, achieving impressive F-score 91.7. novelty this work lies leveraging detection, demonstrating its suitability task. method's ability rapidly continuously segment structural planes without human intervention represents significant advancement, addressing challenges processing image data deep engineering. establishes foundation fully recognition parameter extraction contributing more efficient accurate assessments.

Language: Английский

Integrating inflow control valve control with LSTM networks for oil production forecasting in horizontal intelligent well application DOI Creative Commons
Liang Zhang, Cheng Zhong,

Zhouzheng Hao

et al.

Journal of Petroleum Exploration and Production Technology, Journal Year: 2025, Volume and Issue: 15(3)

Published: Feb. 25, 2025

Language: Английский

Citations

0
Accurate and automated detection of fractures in borehole image using YOLO-V8 model DOI
Yiming Yang, Qingchun Li, Hui Li

et al.

Petroleum Science and Technology, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 19

Published: Nov. 19, 2024

The digital panoramic drilling camera system has produced numerous high-precision borehole images essential for analyzing subsurface geological structures. However, manual segmentation and preprocessing of these are labor-intensive subjective. This paper introduces an innovative approach using the YOLOv8 model, trained on a substantial dataset fractures, to achieve fast, stable, automated recognition. model effectively learns extracts fracture features, accurately identifying defects even in complex contexts such as developed rock cleavage, schistosity, quartz vein bands, mixed alteration. Experimental results indicate that while is slightly slower inference speed than YOLOv5, it excels pixel accuracy, recall rate, mean Average Precision (mAP) segmentation, achieving impressive F-score 91.7. novelty this work lies leveraging detection, demonstrating its suitability task. method's ability rapidly continuously segment structural planes without human intervention represents significant advancement, addressing challenges processing image data deep engineering. establishes foundation fully recognition parameter extraction contributing more efficient accurate assessments.

Language: Английский

Citations

1