A hybrid local-global feature attention network for thin section rock image classification

Key Points

• Local-global feature attention networks enhance classification accuracy for thin section images, advancing geological analysis.
• In experiments, the hybrid approach achieved a 15% increase in classification accuracy compared to traditional methods.
• The study employs a neural network architecture tailored for image classification, integrating both local and global features effectively.
• These findings suggest that improving image classification methods can better support geological research and practical applications.