Texture Analysis to Enhance Drone-Based Multi-Modal Inspection of Structures

The drone-based multi-modal inspection of industrial structures is a relatively new field research gaining interest among companies. Multi-modal can significantly enhance data analysis and provide more accurate assessment the components’ operability structural integrity, which assist in avoiding misinterpretation providing comprehensive evaluation, one NDT4.0 objectives. This paper investigates use coupled thermal visible images to abnormality detection accuracy inspections. Four cases are presented, introducing novel process pipelines for enhancing defect different scenarios. first case presents pipeline feature visibility on using pavement crack detection. second proposes an classification method surface subsurface defects both modalities texture segmentation piping third introduces road modalities. A proposed extract regions images. Further, combination used detect defects. approach employed bridge fourth concrete surfaces