Intelligent identification of pavement cracks based on PSA-Net

The identification of pavement cracks is critical for ensuring road safety. Currently, manual crack detection quite time-consuming. To address this issue, automated crack-detecting technology required. However, automatic recognition remains challenging, owing to the intensity heterogeneity and complexity backdrop, e.g., low contrast damages backdrop may have shadows with comparable intensity. Motivated by breakthroughs in deep learning, we present a new network architecture combining feature pyramid attention mechanism (PSA-Net). In pyramid, integrates spatial information underlying features detection. During training process, it improves accuracy nested sample weighting equalize loss caused simple complex samples. verify effectiveness suggested technique, used dataset real test different methods.