Rock fracture mesoscopic image intelligent identification method and system
By acquiring detailed images of rock fracture surfaces on mobile devices and training a model with an EfficientNetB4 base layer and a custom top-level classifier, the problem of low efficiency and accuracy in traditional rock identification methods is solved, achieving efficient rock identification.
CN121661621APending Publication Date: 2026-03-13CCCC SECOND HARBOR CONSULTANTS CO LTD
Patent Information
- Application Number
- CN202511606020.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-05
- Publication Date
- 2026-03-13
Smart Images

Figure CN121661621A_ABST
Abstract
The invention relates to a rock fracture mesoscopic image intelligent identification method and system, and belongs to the technical field of engineering geological prospecting, and the method comprises the steps: collecting various types of rock fracture mesoscopic images through a mobile device, carrying out the lithologic marking of the rock fracture mesoscopic images, and constructing a rock fracture mesoscopic image data set, the lithology identification method comprises the following steps: providing a high-quality sample for a lithology identification model, taking a rock fracture mesoscopic image data set as training data, training the constructed lithology identification model by adopting a staged training strategy to obtain a trained lithology identification model, the lithology identification model comprising an OfficientNetB4 base layer and a user-defined top layer classifier, and identifying the to-be-identified rock fracture mesoscopic image based on the trained lithology identification model to obtain the lithology identification result of the rock, thereby improving the efficiency and accuracy of geological field work lithology identification.
Need to check novelty before this filing date? Find Prior Art
Citation Information
Patent Citations
Mineral fracture show system and method
CN105335023A
Rock lithology identification method based on NAS, terminal device and storage medium
CN114170589A
Rock slice image multi-stage intelligent identification method
CN116563661A
Intelligent rock image lithology identification method and system based on Inception-v3 transfer learning model
CN119131555A
Lithology intelligent identification method based on median-enhanced attention mechanism
CN119741533A