Artificial reefs are man-made structures submerged in the ocean, and the design of these structures plays a crucial role in determining their effectiveness. Precisely measuring the configuration of artificial reefs is vital for creating suitable habitats for marine organisms. This study presents a novel approach for automated detection of artificial reefs by recognizing their key features and key points. Two enhanced models, namely, YOLOv8n-PoseRFSA and YOLOv8n-PoseMSA, are introduced based on the YOLOv8n-Pose architecture. The YOLOv8n-PoseRFSA model exhibits a 2.3% increase in accuracy in pinpointing target key points compared to the baseline YOLOv8n-Pose model, showcasing notable enhancements in recall rate, mean average precision (mAP), and other evaluation metrics. In response to the demand for swift identification in mobile fishing scenarios, a YOLOv8n-PoseMSA model is proposed, leveraging MobileNetV3 to replace the backbone network structure. This model reduces the computational burden to 33% of the original model while preserving recognition accuracy and minimizing the accuracy drop. The methodology outlined in this research enables real-time monitoring of artificial reef deployments, allowing for the precise quantification of their structural characteristics, thereby significantly enhancing monitoring efficiency and convenience. By better assessing the layout of artificial reefs and their ecological impact, this approach offers valuable data support for the future planning and implementation of reef projects.

Artificial reefs (ARS) are structures placed in the ocean, and their posture and position affect the placement effect of the reef. Quantitative data on the reef posture and position can provide references for the creation of a favorable environment for marine organisms. This paper focuses on improving the quality assessment methods for the deployment of ARS in marine habitats. Based on the images of ARS detected by forward-looking sonar, a new method is proposed to determine the three-dimensional position and deployment effectiveness of the reefs, with the aim of obtaining more detailed information about the ARS to assist in evaluating the quality of their deployment. By constructing a multi-object tracking (MOT) dataset based on the Oculus sonar, the YOLOv8n-pose series model was used for object detection. Subsequently, the three-dimensional position and subsidence of ARS were evaluated using trajectories identified by the MOT algorithm. This paper compares various MOT algorithms and improves the BOTSORT algorithm by introducing the Hungarian algorithm and the Joseph form of the Kalman filter, significantly enhancing the accuracy of target tracking. The experimental results demonstrate that, within the framework of the YOLOv8n-pose model, the detection and association accuracy of the BOTSORT algorithm, which has been refined to address the characteristics of sonar images, has been further enhanced. Additionally, this paper proposes a mathematical modeling method for the assessment of ARS and their subsidence, providing important technical support and solutions for future marine ranch construction.

In recent years, with the development of pelagic fishing, the working environment and monitoring of crew (squid jigger) members have become increasingly important. However, traditional methods of pelagic human observers suffer from high costs, low coverage, poor timeliness, and susceptibility to subjective factors. In contrast, the Electronic Monitoring System (EMS) has advantages such as continuous operation under various weather conditions; more objective, transparent, and efficient data; and less interference with fishing operations. This paper shows how the 3DCNN model, LSTM+ResNet model, and TimeSformer model are applied to video-classification tasks, and for the first time, they are applied to an EMS. In addition, this paper tests and compares the application effects of the three models on video classification, and discusses the advantages and challenges of using them for video recognition. Through experiments, we obtained the accuracy and relevant indicators of video recognition using different models. The research results show that when NUM_FRAMES is set to 8, the LSTM+ResNet-50 model has the best performance, with an accuracy of 88.47%, an $F1$ score of 0.8881, and an $m_{ap}$ score of 0.8133. Analyzing the EMS for pelagic fishing can improve China’s performance level and management efficiency in pelagic fishing, and promote the development of the fishery knowledge service system and smart fishery engineering.

In order to solve the statistical problem of some targets when fishing vessels are operating, based on deep learning technology, this paper uses the improved YOLOv5s and SORT algorithms to achieve target statistics. First, YOLOv5s is fused with CBAM and SE attention mechanism modules, respectively, to reduce the interference of complex backgrounds and improve the model detection accuracy simultaneously. Comparing the three models, the target detection model with a better effect is selected. Secondly, through the threshold method, SORT algorithm combined with the detection line and DeepSORT three algorithms to achieve the count of some targets, respectively. The results show that the accuracies of YOLOv5s, YOLOv5s fused CBAM, and YOLOv5s fused SE are 97.2%, 84.8%, and 98.9%, respectively. Among them, the YOLOv5s fusion SE module has the best effect, which is 1.7% and 14.1% higher than the other two results. Among the three target statistics methods, the SORT algorithm combined with the detection line is the best, with an average count accuracy rate of 85.7%. The count accuracy rates of the three categories of Fish_basket, Fish_net, and Process_ship are 96.5%, 85.8%, and 75%, respectively, compared with the other two species have improved significantly. The research results can provide an informational reference for the automated counting of targets during fishing vessel operations.