Search

Satellite remote sensing technology can achieve real-time observation of ships at sea, and the remote sensing images obtained have the advantages of high contrast and low noise and have become one of the important means of marine monitoring. For the satellite remote sensing image data, there are two main problems: first, remote sensing data class-imbalance problem, and second the existing target detector in the presence of clouds, islands, farmed nets, and other interferences on the small-target ship, and there is a leakage of detection and wrong detection problem. To address the above problems, first, a new dataset containing 3881 images of remotely sensed ships in a variety of complex environments is constructed, which contains a total of 8418 ship instances. Second, we propose CSDP-YOLO for the small-target ship detection method with remote sensing data class imbalance. In order to enhance the performance of neural networks for small-target ship detection in remote sensing images, the innovative CSDP module is proposed, which uses deep large kernel convolution to enhance the sensory field of shallow features and mixes the channel positions using point convolution to obtain a more excellent feature extraction performance. Finally, the MPDIoU loss function is introduced to solve the class-imbalance problem between remote sensing small target ships and the background. We compare the performance with other state-of-the-art algorithms. The experimental results show that the proposed CSDP-YOLO algorithm can significantly improve the performance of small-target ship detection for private datasets. Its average precision, recall, and $\text{AP}_{50}$ are improved to 90.1%, 86.6%, and 91.4%, respectively. For the SSDD public remote sensing dataset, its metrics can reach the highest 93.6%, 93.7%, and 96.8%, respectively.

The normal growth of fishes is closely relevant to the density of mariculture. It is of great significance to accurately calculate the breeding area of specific sea area from satellite remote sensing images. However, there are no reports about cage segmentation and density detection based on remote sensing images so far. And the accurate segmentation of cages faces challenges from very large high-resolution images. Firstly, a new public mariculture cage data set is built. Secondly, the training set is augmented via sample variations to improve the robustness of the model. Then, for cage segmentation and density statistics, a new methodology based on Mask R-CNN is proposed. Using dividing and stitching technologies, the entire remote sensing test images of the cage can be accurately segmented. Finally, using the trained model, the object detection features and segmentation characteristics can be obtained at the same time. Considering only the area within the target detection frame, the proposed method can count the pixels in the segmented area, which can obtain accurate area and density while reducing time-consuming. Experimental results demonstrate that, compared with traditional contour extraction method and U-Net based scheme, the proposed scheme can significantly improve segmentation precision and model’s robustness. The relative error of the actual area is only 1.3%.

Major pests of corn insects include corn borer, armyworm, bollworm, aphid, and corn leaf mites. Timely and accurate detection of these pests is crucial for effective pests control and scientific decision making. However, existing methods for identification based on traditional machine learning and neural networks are limited by high model training costs and low recognition accuracy. To address these problems, we proposed a YOLOv7 maize pests identification method incorporating the Adan optimizer. First, we selected three major corn pests, corn borer, armyworm and bollworm as research objects. Then, we collected and constructed a corn pests dataset by using data augmentation to address the problem of scarce corn pests data. Second, we chose the YOLOv7 network as the detection model, and we proposed to replace the original optimizer of YOLOv7 with the Adan optimizer for its high computational cost. The Adan optimizer can efficiently sense the surrounding gradient information in advance, allowing the model to escape sharp local minima. Thus, the robustness and accuracy of the model can be improved while significantly reducing the computing power. Finally, we did ablation experiments and compared the experiments with traditional methods and other common object detection networks. Theoretical analysis and experimental result show that the model incorporating with Adan optimizer only requires 1/2-2/3 of the computing power of the original network to obtain performance beyond that of the original network. The mAP@[.5:.95] (mean Average Precision) of the improved network reaches 96.69% and the precision reaches 99.95%. Meanwhile, the mAP@[.5:.95] was improved by 2.79%-11.83% compared to the original YOLOv7 and 41.98%-60.61% compared to other common object detection models. In complex natural scenes, our proposed method is not only time-efficient and has higher recognition accuracy, reaching the level of SOTA.

Abstract It is an open issue to train effective deep network models on class imbalance datasets. In the widely used cost‐sensitive imbalanced learning methods, the costs are based on the losses or class probabilities of samples. In this paper, it is discovered that these traditional cost‐sensitive methods discard the clustering feature, and introduce the errors of annotations into costs, leading to sub‐optimal models. It is further investigated that the feature magnitude of sample, which is computed before probability and loss, not only is independent of the annotation, but also represents the familiarity degree of model with the sample. These characteristics of feature magnitude are used to guide the training and inference of model. First, the concept of stranger is proposed, which is the sample with small feature magnitude value, and the idea of focal learning on strangers (FLS) is proposed. By adding the idea of FLS into two existing cost‐sensitive methods, two novel losses are put forward: instance‐level focal stranger loss (IFSL) and class‐level focal stranger loss (CFSL). The losses can improve the aggregation features of samples within class, and reduce the negative influences of annotation errors on imbalanced learning. Second, considering the large difference of feature magnitude means between minority class and majority class in case of extreme class‐imbalance dataset, a bias determination (BD) strategy is put forward to improve classification performance during inference. The methods are applied to the tasks of image semantic segmentation and salient‐instance segmentation. The experimental results on four public semantic segmentation datasets demonstrate that IFSL can reduce the over‐fitting of model, improve the classification accuracy of rare samples, and alleviate the reliance of performance on the annotation quality. The experimental results on two public salient‐instance segmentation datasets show that CFSL makes the model have better scoring ability for salient object. Besides, the BD strategy can reduce the wrong classification caused by bias model. Therefore, the proposed methods can significantly advance image segmentation.

Abstract Imbalanced datasets commonly exist in real world and greatly challenge the performances of deep neural models. The authors discover that the traditional balance strategies in the existing imbalanced learning methods emphasize/suppress the importance of minority/majority class in a fixed way. Even the minority class is fully represented, the minority/majority class is still emphasized/suppressed, eventually leading to sub‐optimal models. The authors argue that the importance of minority/majority class should be emphasized/suppressed in a dynamic manner. They propose to dynamically balance class losses (DBCL), which first suppresses the importance of majority class, and then gradually restores the importance of majority class. By adding the idea of DBCL into three state‐of‐the‐art cost‐sensitive loss functions, the authors propose three novel loss functions: dynamic constrained‐focal‐loss (DCFL), dynamic balance‐cross‐entropy based on the effective number of samples (DBENS) and dynamic balance‐cross‐entropy based on class frequencies (DBCEF). They apply DCFL, DBENS and DBCEF to the imbalanced learning in salient‐instance segmentation. The results show that their methods significantly improve the classification abilities of models.

In wireless communication, the channel function estimated commonly has errors due to the influence of noise, so traditional channel estimation methods cannot accurately estimate the real channel function. Aiming at this problem, we propose a channel estimation method that combines sounding reference signal (SRS) remapping with the deep-learning network BRD_ESRNet. BRD_ESRNet consists of image denoising using a deep convolutional neural network with batch renormalization (BRDNet) and an expanded superresolution convolutional neural network (ESRCNN). At the transmitter side, we first map the SRS into four-box structures, and then, the four-box structures are scattered distribution throughout the time-frequency resource block. At the receiver side, we first perform the modified least squares (LS) estimation based on the four-box structure and place the result into the top-left resource unit of the four box. Then, we perform linear interpolation for the whole resource block. Finally, we equate the estimated channel matrix to a low-resolution image containing noise and input it to BRD_ESRNet. Thus, we obtain data with high resolution and achieve the purpose of reducing the estimation error of the channel function. The experimental results show that the proposed method in this paper has a significant improvement in performance compared to the methods of Soltani et al. and Nithya et al. In this paper, the methods of Soltani et al. and Nithya et al. are referred to as methods 1 and 2, respectively.

In clinical drug sensitivity experiments, it is necessary to plate culture pathogenic bacteria and pick suitable colonies for bacterial solution preparation, which is a process that is currently carried out completely by hand. Moreover, the problems of plate contamination, a long culture period, and large image annotation in colony plate image acquisition can lead to a small amount of usable data. To address the issues mentioned above, we adopt a deep learning approach and conduct experiments on the AGAR dataset. We propose to use style transfer to extend the trainable dataset and successfully obtain 4k microbial colony images using this method. In addition, we introduce the Swin Transformer as a feature extraction network in the Cascade Mask R-CNN model architecture to better extract the feature information of the images. After our experimental comparison, the model achieves a mean Average Precision (mAP) of 61.4% at the Intersection over Union (IoU) [0.50:0.95]. This performance surpasses that of the Cascade R-CNN with HRNet, which is the top-performing model in experiments conducted on the AGAR dataset, by a margin of 2.2%. Furthermore, we perform experiments using YOLOv8x on the AGAR dataset, which results in a mAP of 76.7%.

In this paper, an optical fiber humidity sensor based on acrylate AB adhesive film is studied, and two methods—a bubbling method and a dual pressure assisted method—for preparing thin films are proposed. The forms will make the thin film lighter and make the sensor more sensitive. Using a glass tube to connect the acrylate AB adhesive film to the single mode optical fiber, the humidity sensor is obtained. Through humidity response experiments, the sensor’s sensitivity reaches 167.5 pm/% RH, and the response time reaches 4.8 s/% RH. At the same time, the experiments show that the sensor has good repeatability and stability. Finally, the influence of temperature on the working process is analyzed, and we provide a method for improvement through FBG.

The morphological features of fish, such as the body length, the body width, the caudal peduncle length, the caudal peduncle width, the pupil diameter, and the eye diameter are very important indicators in smart mariculture. Therefore, the accurate measurement of the morphological features is of great significance. However, the existing measurement methods mainly rely on manual measurement, which is operationally complex, low efficiency, and high subjectivity. To address these issues, this paper proposes a scheme for segmenting fish image and measuring fish morphological features indicators based on Mask R-CNN. Firstly, the fish body images are acquired by a home-made image acquisition device. Then, the fish images are preprocessed and labeled, and fed into the Mask R-CNN for training. Finally, the trained model is used to segment fish image, thus the morphological features indicators of the fish can be obtained. The experimental results demonstrate that the proposed scheme can segment the fish body in pure and complex backgrounds with remarkable performance. In pure background, the average relative errors (AREs) of all indicators measured all are less than 2.8%, and the AREs of body length and body width are less than 0.8%. In complex background, the AREs of all indicators are less than 3%, and the AREs of body length and body width is less than 1.8%.

In the smart mariculture, the timely and accurate predictions of water quality can help farmers take countermeasures before the ecological environment deteriorates seriously. However, the openness of the mariculture environment makes the variation of water quality nonlinear, dynamic and complex. Traditional methods face challenges in prediction accuracy and generalization performance. To address these problems, an accurate water quality prediction scheme is proposed for pH, water temperature and dissolved oxygen. First, we construct a new huge raw data set collected in time series consisting of 23,204 groups of data. Then, the water quality parameters are preprocessed for data cleaning successively through threshold processing, mean proximity method, wavelet filter, and improved smoothing method. Next, the correlation between the water quality to be predicted and other dynamics parameters is revealed by the Pearson correlation coefficient method. Meanwhile, the data for training is weighted by the discovered correlation coefficients. Finally, by adding a backward SRU node to the training sequence, which can be integrated into the future context information, the deep Bi-S-SRU (Bi-directional Stacked Simple Recurrent Unit) learning network is proposed. After training, the prediction model can be obtained. The experimental results demonstrate that our proposed prediction method achieve higher prediction accuracy than the method based on RNN (Recurrent Neural Network) or LSTM (Long Short-Term Memory) with similar or less time computing complexity. In our experiments, the proposed method takes 12.5ms to predict data on average, and the prediction accuracy can reach 94.42% in the next 3~8 days.

In harbour areas, the irrational layout and high density of mariculture cages can lead to a dramatic deterioration of the culture’s ecology. Therefore, it is important to analyze and regulate the distribution of cages using intelligent analysis based on deep learning. We propose a remote sensing image segmentation method based on the Swin Transformer and ensemble learning strategy. Firstly, we collect multiple remote sensing images of cages and annotate them, while using data expansion techniques to construct a remote sensing image dataset of mariculture cages. Secondly, the Swin Transformer is used as the backbone network to extract the remote sensing image features of cages. A strategy of alternating the local attention module and the global attention module is used for model training, which has the benefit of reducing the attention computation while exchanging global information. Then, the ensemble learning strategy is used to improve the accuracy of remote sensing cage segmentation. We carry out quantitative and qualitative analyses of remote sensing image segmentation of cages at the ports of Li’an, Xincun and Potou in Hainan Province, China. The results show that our proposed segmentation scheme has significant performance improvement compared to other models. In particular, the mIoU reaches 82.34% and pixel accuracy reaches 99.71%.

The statistics of disease spores is significant for early strategy design of disease control in precision agriculture. To obtain the statistics information of spores in microscopic images, it is crucial to segment spores from images. In this paper, we research a deep learning based method to segment spores, taking anthrax spores as the research objects. We first built an anthrax spore dataset consisting of more than 40,000 spores with accurate labeled spore boundaries to advance the state of the art technology of spore statistics. Then on consideration of the complex class imbalances in actual anthrax spore images, we investigate how class imbalances and hard examples simultaneously influence the loss during training and we discover that hard examples are more likely to appear at the pixels of rare pixels, such as small class pixels and contour pixels. Based on this discovery, we propose Constrained Focal Loss (CFL), which focuses on small class objects, and has a constrained term related to hard examples. In addition, we further propose CFL*, where high importance is put on the pixels surrounding spore contours to improve classification accuracy. The results show that the mean IoU of the DeepLabv3+ trained with CFL* (called as CFL*Net) achieves 91.0%, higher than original DeepLabv3+ with cross-entropy by 8.6 points, and the DeepLabv3+ with Focal Loss by 10.4 points. Moreover, CFLNet* can achieve better performance than original DeepLabv3+, using less than one-third of the training samples and half of the training steps.

Spectrum sensing is one of the key technologies in wireless wideband communication. There are still challenges in respect of how to realize fast and robust wideband spectrum sensing technology. In this paper, a novel nonreconstructed sequential compressed wideband spectrum sensing algorithm (NSCWSS) is proposed. Firstly, the algorithm uses a sequential spectrum sensing method based on history memory and reputation to ensure the robustness of the algorithm. Secondly, the algorithm uses the strategy of compressed sensing without reconstruction, which thus ensures the sensing agility of the algorithm. The algorithm is simulated and analyzed by using the centralized cooperative sensing. The theoretical analysis and simulation results reveal that, under the condition of ensuring the certain detection probability, the proposed algorithm effectively reduces complex computation of signal reconstruction, significantly reducing the wideband spectrum sampling rate. At the same time, in the cognitive wideband communication scenarios, the algorithm also achieves a better defense against the SSDF attack in spectrum sensing.

An accurate prediction of cage-cultured water quality is a hot topic in smart mariculture. Since the mariculturing environment is always open to its surroundings, the changes in water quality parameters are normally nonlinear, dynamic, changeable, and complex. However, traditional forecasting methods have lots of problems, such as low accuracy, poor generalization, and high time complexity. In order to solve these shortcomings, a novel water quality prediction method based on the deep LSTM (long short-term memory) learning network is proposed to predict pH and water temperature. Firstly, linear interpolation, smoothing, and moving average filtering techniques are used to repair, correct, and de-noise water quality data, respectively. Secondly, Pearson’s correlation coefficient is used to obtain the correlation priors between pH, water temperature, and other water quality parameters. Finally, a water quality prediction model based on LSTM is constructed using the preprocessed data and its correlation information. Experimental results show that, in the short-term prediction, the prediction accuracy of pH and water temperature can reach 98.56% and 98.97%, and the time cost of the predictions is 0.273 s and 0.257 s, respectively. In the long-term prediction, the prediction accuracy of pH and water temperature can reach 95.76% and 96.88%, respectively.

In order to enhance the enthusiasm of the data provider in the process of data interaction and improve the adequacy of data interaction, we put forward the concept of the ego of data and then analyzed the characteristics of the ego of data in the Internet of Things (IOT) in this paper. We implement two steps of data clustering for the Internet of things; the first step is the spatial location of adjacent fuzzy clustering, and the second step is the sampling time fuzzy clustering. Equivalent classes can be obtained through the two steps. In this way we can make the data with layout characteristics to be classified into different equivalent classes, so that the specific location information of the data can be obscured, the layout characteristics of tags are eliminated, and ultimately anonymization protection would be achieved. The experimental results show that the proposed algorithm can greatly improve the efficiency of protection of the data in the interaction with others in the incompletely open manner, without reducing the quality of anonymization and enhancing the information loss. The anonymization data set generated by this method has better data availability, and this algorithm can effectively improve the security of data exchange.

Privacy-preserving in wireless sensor networks is one of the key problems to be solved in practical applications. It is of great significance to solve the problem of data privacy protection for large-scale applications of wireless sensor networks. The characteristics of wireless sensor networks make data privacy protection technology face serious challenges. At present, the technology of data privacy protection in wireless sensor networks has become a hot research topic, mainly for data aggregation, data query, and access control of data privacy protection. In this paper, multiorder fusion data privacy-preserving scheme (MOFDAP) is proposed. Random interference code, random decomposition of function library, and cryptographic vector are introduced for our proposed scheme. In multiple stages and multiple aspects, the difficulty of cracking and crack costs are increased. The simulation results demonstrate that, compared with the typical Slice-Mix-AggRegaTe (SMART) algorithm, the algorithm proposed in this paper has a better data privacy-preserving ability when the traffic load is not very heavy.

The sampling rate of wideband spectrum sensing for sparse signals can be reduced by sub-Nyquist sampling with a Modulated Wideband Converter (MWC). In collaborative spectrum sensing, the fusion center recovers the spectral support from observation and measurement matrices reported by a network of CRs, to improve the precision of spectrum sensing. However, the MWC has a very high hardware complexity due to its parallel structure; it sets a fixed threshold for a decision without considering the impact of noise intensity, and needs a priori information of signal sparsity order for signal support recovery. To address these shortcomings, we propose a progressive support selection based self-adaptive distributed MWC sensing scheme (PSS-SaDMWC). In the proposed scheme, the parallel hardware sensing channels are scattered on secondary users (SUs), and the PSS-SaDMWC scheme takes sparsity order estimation, noise intensity, and transmission loss into account in the fusion center. More importantly, the proposed scheme uses a support selection strategy based on a progressive operation to reduce missed detection probability under low SNR levels. Numerical simulations demonstrate that, compared with the traditional support selection schemes, our proposed scheme can achieve a higher support recovery success rate, lower sampling rate, and stronger time-varying support recovery ability without increasing hardware complexity.