基于计算机视觉的预制小箱梁外观智能评估方法和系统
By using the Transformer-based image segmentation model SegFormer and morphological operations, the subjectivity of manual inspection and the adaptability of computer vision methods in the appearance evaluation of precast small box girders are solved, enabling accurate quantification and evaluation of defects of different shapes, and improving the objectivity and efficiency of the evaluation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CCCC HIGHWAY BRIDGES NATIONAL ENGINEERING RESEARCH CENTRE CO LTD
- Filing Date
- 2025-09-19
- Publication Date
- 2026-07-17
AI Technical Summary
In existing technologies, the appearance evaluation of precast small box girders relies on manual inspection, which results in highly subjective results, low efficiency, and easy to miss subtle defects. Existing computer vision methods are difficult to adapt to the characteristics of defects of different shapes, are prone to losing key information, and cannot balance noise reduction and feature preservation.
Defect identification is performed using the Transformer-based image segmentation model SegFormer. Combined with 8-neighborhood connectivity analysis and morphological operations, the defect region is cleaned up through erosion-dilation processing, the geometric size of the defect is quantified, and accurate evaluation is achieved through pixel-to-physical size conversion coefficient.
It enables accurate segmentation and quantification of appearance defects in precast small box girders, improving the objectivity, accuracy, and efficiency of the assessment, and supporting quality control in high-quality engineering construction.
Smart Images
Figure CN121353177B_ABST