Zhongwei Xu

Underwater pile foundation detection is crucial for environmental monitoring and marine engineering. Traditional methods for detecting underwater pile foundations are labor-intensive and inefficient. Deep learning-based image processing has revolutionized detection, enabling identification through sonar imagery analysis. This study proposes an innovative methodology, named the AquaPile-YOLO algorithm, for underwater pile foundation detection. Our approach significantly enhances detection accuracy and robustness by integrating multi-scale feature fusion, improved attention mechanisms, and advanced data augmentation techniques. Trained on 4000 sonar images, the model excels in delineating pile structures and effectively identifying underwater targets. Experimental data show that the model can achieve good target identification results in similar experimental scenarios, with a 96.89% accuracy rate for underwater target recognition.

In recent years, interest in aquaculture acoustic signal has risen since the development of precision agriculture technology. Underwater acoustic signals are known to be noisy, especially as they are inevitably mixed with a large amount of environmental background noise, causing severe interference in the extraction of signal features and the revelation of internal laws. Furthermore, interference adds a considerable burden on the transmission, storage, and processing of data. A signal recognition curve (SRC) algorithm is proposed based on higher-order cumulants (HOC) and a recognition-sigmoid function for feature extraction of target signals. The signal data of interest can be accurately identified using the SRC. The analysis and verification of the algorithm are carried out in this study. The results show that when the SNR is greater than 7 dB, the SRC algorithm is effective, and the performance improvement is maximized when the SNR is 11 dB. Furthermore, the SRC algorithm has shown better flexibility and robustness in application.