一种高光谱岩性识别巷道建模方法
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUIZHOU INST OF TECH
- Filing Date
- 2025-11-12
- Publication Date
- 2026-07-17
AI Technical Summary
Existing hyperspectral systems are bulky and difficult to disassemble, with fixed field of view for imaging spectral units, resulting in poor versatility and difficulty in efficiently identifying roadway lithology in outdoor scenarios.
A mobile hyperspectral-geometric synchronous acquisition system was adopted, which combined tightly coupled synchronous positioning and map building algorithms to acquire hyperspectral images, depth point clouds and inertial measurement unit data in real time. Data was repaired by Kriging interpolation, and a two-stream three-dimensional convolutional neural network was applied for lithological classification and semantic constraint registration to generate a three-dimensional semantic tunnel model.
It has achieved automation and precision in roadway lithology identification, improved scanning efficiency by more than 20 times, reduced manual intervention, and generated models with accurate geometric morphology and real lithological properties. It supports the automatic calculation of rock quality indicators and expands the practical value of the models.
Smart Images
Figure CN121482573B_ABST