Accurate Water Gauge Detection by Image Data Augmentation Using Erase-Copy-Paste

Water level recognition is significant for flood warnings in the hydrology industry, which mainly includes water gauge detection and water level line recognition. With the video surveillance system becoming the basic facility of hydrological stations, computer vision technologies are widely used for water gauge detection. However, it is commonly observed that the lack of water gauge images critically impacts the performance of deep learning. To address this issue to achieve accurate water gauge detection, this study proposes a data augmentation strategy called ECP. This strategy tackles the shortage of water gauge images by mixing them, thereby expanding the dataset and ultimately enhancing the accuracy of water gauge detection. We conducted two experiments to evaluate the effectiveness of ECP using a practical hydrology dataset obtained from the hydrological station system in Wuyuan, Jiangxi Province. Numerical results show that our proposed ECP can promote average precision and average recall of representative object detection algorithms. In addition, compared with other mixed-image methods, the performance improvement of ECP is better. Our experimental results indicated that the ECP can increase the dataset diversity, and enhance the water gauge detection algorithm accuracy and generalization ability in practical industrial applications.