A weak light equipment inspection method based on adaptive multi-spectrum fusion
By deploying a multispectral acquisition unit in the wind tunnel equipment and performing adaptive decomposition and joint sensing processing, the problem of wind tunnel equipment inspection in low-light environments has been solved, enabling precise anomaly detection and location on the blade surface, and improving inspection efficiency and accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SICHUAN PROVINCIAL IND EQUIP INSTALLATION CO
- Filing Date
- 2026-03-24
- Publication Date
- 2026-06-16
AI Technical Summary
In the dark or poorly lit environment of wind tunnel equipment, traditional equipment inspection methods are unable to obtain clear and complete images of the blade surface, making it impossible to accurately determine whether there are abnormalities such as loosening on the blade surface. Furthermore, existing inspection methods lack effective fusion of multispectral data, making it impossible to comprehensively obtain various information about the blade surface.
By deploying a multispectral acquisition unit in a low-light environment to acquire visible light, near-infrared, and thermal infrared spectral image sequences of the equipment surface, adaptive decomposition processing is performed to decompose the images into spatial structure, spectral material, and thermal radiation dominant component layers. A pre-built spatial-spectral joint sensing and reconstruction network is used to perform cross-component feature association and enhancement processing to generate a multispectral joint feature tensor. Finally, a spatial distribution map of the equipment's abnormal state is generated, and the unmanned inspection aircraft is triggered to perform differentiated spectral mode verification.
It enables precise inspection of wind tunnel equipment blade surfaces in low-light environments, automatically identifies and locates various abnormalities, and improves inspection efficiency and accuracy.
Smart Images

Figure 1 
Figure 2