Entropy-Based Automated Method for Detection and Assessment of Spalling Severities in Reinforced Concrete Bridges

Existing bridges are aging and deteriorating rapidly, elevating concerns for public safety preservation of these valuable assets. Large numbers exist in transportation networks, the current budget limitations worsen situation. This necessitates development an automated condition assessment rating methods. Spalling is a common problem that majorly influences health, safety, structural integrity bridges. The present study introduces self-adaptive three-tier method detection spalling using computer-vision technologies. first model newly-developed segmentation adopts multiobjective invasive weed optimization information theory-based formalism images spalled concrete detection. In second model, integration singular value decomposition discrete wavelet transform integrated efficient feature extraction images. Additionally, Elman neural network coupled with algorithm to enhance accuracy evaluation severities by amplifying exploration-exploitation trade-off mechanism network. third developed purpose structuring system severity based on its area depth. A computerized platform C#.net language facilitate implementation users. results demonstrated capable improving 12.29% respect region growing algorithm. It was also inferred quantification outperformed other prediction models, such it achieved mean absolute percentage error, root mean-squared squared error 4.07%, 76.061, 0.065, respectively, original dataset. this regard, expected computer-vision-based can aid establishing cost-effective bridge models agencies.