In response to the challenges of low detection accuracy, missed detections, and false detections in identifying valuable marine organisms in marine ranch areas, this study introduces an enhanced algorithm for detecting rare underwater marine species using the YOLOv8 model. Firstly, a new residual attention mechanism was designed and integrated into the backbone network of the YOLOv8 model to improve focus on the detailed features of underwater targets during feature extraction. Next, a bidirectional feature pyramid with adaptive feature fusion and feature selection characteristics is incorporated into the neck network to effectively combine the strong semantic information of deep feature maps with the localization information of shallow feature maps. This emphasizes the distinctions between the target and the surroundings. The experiment showed that the mean Average Precision (mAP@0.5) of the enhanced YOLOv8 model was 92.98%, which is 1.36 percentage point higher than the original YOLOv8 model. Additionally, the mean Average Precision (mAP@0.5∶0.95) was 76.71%, indicating a 3.7 percentage point improvement over the original YOLOv8 model. Compared with mainstream object detection models such as Faster RCNN, SSD, RetinaNet, YOLOv6, and YOLOv7, the improved model has shown an increase of 1.57 percentage point, 1.74 percentage point, 3.17 percentage point, 4.68 percentage point, and 1.47 percentage point respectively in mAP@0.5. The model proposed in this paper demonstrates high detection accuracy and robust stability in complex seabed environments. It can provide technical support for the scientific management of underwater resources in marine ranches.

Saline-alkali tolerant rice often grows in saline-alkali land with salt (alkali) concentration above 0.3%, mainly in Inner Mongolia and Ningxia semi-desert inland areas. The growth environment has posed a serious threat to the yield of saline-alkali-tolerant rice. Traditional wireless sensor network (WSN) has been widely used for energy consumption, transmission delay and throughput in data collection and transmission. However, the current environmental monitoring cannot fully meet the requirement of large-scale production in recent years, due to the complex environmental data under the condition of low communication coverage. In this study, the medium access layer saline-alkali tolerant rice environmental data perception strategy (MAC-SREP) was proposed in the WSN network, in order to comprehensively optimize the MAC layer and multiple unmanned aerial vehicles (UAV) route planning in the multiple unmanned aerial vehicles-wireless sensor network (mUAVs-WSN) system. The monitoring region was first divided into some subregions using the concave and convex polygon division. Multi-UAVs collaborative search region model was utilized to map into a single UAV search subregion model. Voronoi diagram was also modified using the communication distance of cluster head nodes and the communication coverage radius of the UAV to the ground. The modified Voronoi diagram was then divided into some clusters to optimize the flight path of the UAV. Then, the MAC layer mechanism was used to prioritize the packet types of UAVs for the effective allocation of time slots and network resources. The competition window size was adjusted under the different network loads. The adaptive retreat mechanism was used to reduce the channel conflict. The environment data was collected from the saline-alkali tolerant rice growth in real time and then uploaded into the server of the management center. The data was also processed and then published on the website. The functions of remote monitoring were realized for the historical record query and visualization of the growth environment information of saline-alkali tolerant rice. The field test was carried out in the large-scale continuous planting area at the saline-alkali tolerant rice experimental base of Guangdong Ocean University, Buchao Village, Jianxin Town, Suixi County, Zhanjiang City, Guangdong Province, China. A monitoring area was about 53 hm². A simulation test was also performed on Matlab software. The results show that: 1) The life cycle of information perception in the mUAVs-WSN network was about 1.25 times that of the single unmanned aerial vehicle-wireless sensor network (sUAV-WSN). The network throughput increased by 15%, whereas, the end-to-end average delay was reduced by 26.60%. 2) In the actual engineering test, the packet loss rate of the mUAVs-WSN network was less than 2% with the data error less than 5%, both of which fully met the monitoring requirements of saline-alkali tolerant rice growing environment area information. 3) The mUAVs-WSN was more suitable for the monitoring areas with a large number of nodes and coverage, especially for the agricultural environments, such as the saline-alkali tolerant rice. The naturally harsh environment limited the formation of a complete monitoring network (often divided into multiple independent networks). Therefore, the improved strategy can be expected to realize the stable real-time acquisition and transmission of saline-alkali tolerant rice remote data. The data packet loss rate and delay were effectively reduced to improve the overall performance of the network. The finding can fully meet the needs of long-term, stable, and efficient growth environment data with the large coverages in the saline-alkali tolerant rice.

Sea cucumber is required to accurately and rapidly detect under the underwater complex environments of ocean Sea cucumber is required to accurately and rapidly detect under the underwater complex environments of ocean ranching. However, the small size of the sea cucumbers target is difficult to distinguish from the background. Particularly, the challenge of detection can also be found under the weak lighting, serious noise, and occlusion due to the overlapping sea cucumbers. Therefore, this study aims to propose the YOLOv10-MECAS model improved by the YOLOv10s baseline to enhance the performance of detection. A median pooling enhanced channel attention and a spatial attention were used to design the MECAS (median-enhanced channel and spatial attention) module. MECAS effectively retained the target features and reduced the image noise. Sea cucumber features were then captured using multiscale depth wise convolution. Additionally, the SAConv (switchable atrous convolution) module was introduced to replace the standard 3×3 convolutional module in the SCDown (spatial-channel decoupling downsampling) module. The receptive field was expanded without increasing the convolution kernel size. Thereby the model was improved to capture the features of occluded targets. An enhancement algorithm was employed on the underwater image using UDCP (the underwater dark channel prior). The dataset images were enhanced to significantly optimize the contrast for the high image quality. Furthermore, the MPDIoU (minimum points distance intersection over union) regression loss function was adopted to reduce the distortion of detection boxes caused by large sample variability. Thereby the robustness of the model was enhanced. An experiment was carried out to evaluate the performance of the improved model. A dataset of sea cucumbers was sampled from real underwater scenarios. The experimental results show that the better performance of the improved model was achieved on the original dataset, with a prediction precision of 85.7%, recall of 81.5%, and the mean average precision at IoU (intersection over union) 0.5 of 89.7%, indicting the improvement by 6.4%, 4.4%, and 5.0% over the baseline model. Compared with the comparison models Faster-RCNN, SSD, YOLOv5s, YOLOv7, YOLOv8s, YOLOv9s and YOLOv11s, while maintaining advantages in terms of the number of parameters, GFLOPs (giga floating-point operations per second), and FPS (frames per seconds), the mean average precision at IoU 0.5 had been improved by 16.5, 15.4, 5.5, 6.1, 5.3, 5.8, and 4.6 percentage points. On the enhanced dataset by UDCP algorithm, the improved model was achieved in a prediction precision of 86.4%, recall of 82.6%, and the mean average precision at IoU 0.5 of 90.4%, indicating the improvement of 3.3%, 2.1%, and 4.8% over the baseline model. Compared with the comparison models Faster-RCNN, SSD, YOLOv5s, YOLOv7, YOLOv8s, YOLOv9s and YOLOv11s, the mean average precision at IoU 0.5 had been improved by 16.1, 15.4, 5.5, 5.9, 5.2, 6.1, and 4.3 percentage points. The MECAS module into the YOLOv10s also outperformed the combination of current mainstream attention modules, such as LSKA (large selective kernel attention), CA (coordinate attention) and ECA (efficient channel attention) in underwater sea cucumber detection. Finally, the experiment verified that the YOLOv10s with the MPDIoU also performed better than that with CIoU (complete intersection over union), EIoU (enhanced intersection over union), GIoU(generalized intersection over union), DIoU (distance intersection over union), and SIoU (scaled intersection over union). Consequently, the detection accuracy of small target sea cucumbers was effectively improved in complex underwater environments. The finding can provide a theoretical basis to detect the sea cucumber during harvesting.

Agricultural sensors can greatly contribute to future technologies and systemic innovation in smart agriculture. However, the types and precision of sensors are limited to monitoring the agricultural environment with complex and diverse objects. The large and redundant monitoring data has also resulted in the low reliability of information perception. In this study, an improved radial basis function neural network (RBFNN) and Chernobyl disaster optimizer (ICDO) multi-sensor data fusion was proposed to improve the accuracy and reliability of single-sensor measurement. Firstly, an improved Chernobyl catastrophe optimization was performed on the neural network model. The good-point set theory was introduced to improve the initial population quality of the CDO, particularly for accuracy and speed. The adaptive Laplacian crossover operator was added to enhance the search performance. The better adaptive behavior was achieved in the high convergence speed. And then, the individual learning and differential evolution strategy were used to redefine the location update equation, in order to balance the local and global exploration. Secondly, the RBF neural network model was optimized by ICDO, in order to improve the stability of the model. Finally, the nonlinear mapping of the RBF neural network model was used to realize the multi-sensor data fusion with high accuracy. Three experiments were conducted to verify the improved model. The first one was to verify the ICDO. A large improvement was obtained in the solution accuracy and optimization stability, compared with particle swarm optimization (PSO), gray wolf optimization (GWO), firefly algorithm (FA), dung beetle optimizer (DBO), and subtraction average-based optimizer (SABO). The second one was to evaluate the quality of the atmospheric environment. Specifically, the atmospheric data was collected outside the South Subtropical Botanical Garden in Mazhang District, Zhanjiang City, Guangdong Province, China, from September 1, 2022, to September 30, 2023. The goodness of fit reached 0.999 for the prediction of atmospheric environmental quality, the mean square error was as low as 0.348, and the mean absolute percentage error was reduced to 0.729%. The third one was to classify the greenhouse environment. The data was collected in the greenhouses of the South Asian Tropical Botanical Garden. The accuracy rate of greenhouse environment classification was 99.21% with a precision rate of 99.91%. The data fusion was suitable for both indoor and outdoor environments, indicating better adaptability and high accuracy. This finding can also provide solid technical support to agricultural sensor data fusion in the field of precision agriculture.