AI-Based Structural Defect Detection from Construction Site Images Using Deep Learning Approaches
Annotatsiya
Structural defect detection in construction sites is crucial for ensuring building safety and preventing catastrophic failures. Traditional manual inspection methods are time-consuming, subjective, and prone to human error. This paper presents a comprehensive deep learning-based framework for automated detection of structural defects including cracks, spalling, corrosion, and delamination from construction site images. We propose an enhanced Convolutional Neural Network (CNN) architecture integrated with attention mechanisms and transfer learning to achieve superior detection accuracy. Our method was evaluated on the SDNET2018 and Bridge Crack Image datasets, achieving 96.8% accuracy, 95.2% precision, and 94.7% recall, outperforming state-of-the-art approaches. The system demonstrates real-time processing capabilities with an average inference time of 42ms per image, making it suitable for deployment in automated construction monitoring systems. Our results indicate that AI-based defect detection can significantly reduce inspection costs by up to 65% while improving detection reliability.
Hali tarjima qilinmagan