Search Results (465)

Deep learning techniques have become the mainstream approach for fine-grained crop classification in unmanned aerial vehicle (UAV) remote sensing imagery. However, a significant challenge lies in the long-tailed distribution of crop samples. This imbalance causes neural networks to focus disproportionately on majority class features during training, leading to biased decision boundaries and weakening model performance. We designed a minority sample enhanced sampling (MES) method with the goal of addressing the performance limitations that are caused by class imbalance in many crop classification models. The main principle of MES is to relate the re-sampling probability of each class to the sample pixel frequency, thereby achieving intensive re-sampling of minority classes and balancing the training sample distribution. Meanwhile, during re-sampling, data augmentation is performed on the sampled images to improve the generalization. MES is simple to implement, is highly adaptable, and can serve as a general-purpose sampler for semantic segmentation tasks, functioning as a plug-and-play component within network models. To validate the applicability of MES, experiments were conducted on four classic semantic segmentation networks. The results showed that MES achieved mIoU improvements of +1.54%, +4.14%, +2.44%, and +7.08% on the Dali dataset and +2.36%, +0.86%, +4.26%, and +2.75% on the Barley Remote Sensing Dataset compared with the respective benchmark models. Additionally, our hyperparameter sensitivity analysis confirmed the stability and reliability of the method. MES mitigates the impact of class imbalance on network performance, which facilitates the practical application of deep learning in fine-grained crop classification. Full article

Anomaly detection (AD) is an important topic in remote sensing, aiming to identify unusual or abnormal features within the data. However, most existing low-rank representation methods usually use the nuclear norm for background estimation, and do not consider the different contributions of different singular values. Besides, they overlook the spatial relationships of abnormal regions, particularly failing to fully leverage the 3D structured information of the data. Moreover, noise in practical scenarios can disrupt the low-rank structure of the background, making it challenging to separate anomaly from the background and ultimately reducing detection accuracy. To address these challenges, this paper proposes a weighted multidirectional sparsity regularized low-rank tensor representation method (WMS-LRTR) for AD. WMS-LRTR uses the weighted tensor nuclear norm for background estimation to characterize the low-rank property of the background. Considering the correlation between abnormal pixels across different dimensions, the proposed method introduces a novel weighted multidirectional sparsity (WMS) by unfolding anomaly into multimodal to better exploit the sparsity of the anomaly. In order to improve the robustness of AD, we further embed a user-friendly plug-and-play (PnP) denoising prior to optimize the background modeling under low-rank structure and facilitate the separation of sparse anomalous regions. Furthermore, an effective iterative algorithm using alternate direction method of multipliers (ADMM) is introduced, whose subproblems can be solved quickly by fast solvers or have closed-form solutions. Numerical experiments on various datasets show that WMS-LRTR outperforms state-of-the-art AD methods, demonstrating its better detection ability. Full article

Analyzing and mining spatiotemporal processes refers to the extraction of geographic phenomena from spatiotemporal data and the analysis of available geographic knowledge and patterns. It finds applications in various fields such as natural disaster evolution, environmental pollution, and human behavior prediction. However, training spatiotemporal models based on big data is time-consuming, and the traditional physical models and static objects used in existing geographic data analysis software have limitations in mining efficiency and simulation accuracy for dynamic spatiotemporal processes. In this paper, we develop an intelligent spatiotemporal process analysis and mining software tool, called STPam, which integrates a plug-and-play artificial intelligence model by a service-oriented method, distributed deep learning framework, and multi-source big data adaptation. The floods in the middle reaches of the Yangtze River have perennially affected safety and property in surrounding cities and communities. Therefore, this article applies the software to simulate the flooding process in the basin in 2022. The experimental results correspond to the rare drought phenomenon in the basin, demonstrating the practicality of the STPam software. In summary, STPam aids researchers in visualizing and analyzing geospatial processes and also holds potential application value in assisting regional management authorities in making disaster prevention and mitigation decisions. Full article

The walnut (Juglans regia) is a woody oilseed crop with high economic and food value as its kernels are edible and its hulls can be widely used in oil extraction and plugging, chemical raw materials, and water purification. Currently, red walnut varieties have emerged, attracting consumer interest due to their high nutritional values as they are rich in anthocyanins. WD40 is a widespread superfamily in eukaryotes that play roles in plant color regulation and resistance to stresses. In order to screen for JrWD40 associated with walnut color, we identified 265 JrWD40s in walnuts by genome-wide identification, which were unevenly distributed on 16 chromosomes. According to the phylogenetic tree, all JrWD40s were classified into six clades. WGD (Whole genome duplication) is the main reason for the expansion of the JrWD40 gene family. JrWD40s were relatively conserved during evolution, but their gene structures were highly varied; lower sequence similarity may be the main reason for the functional diversity of JrWD40s. Some JrWD40s were highly expressed only in red or green walnuts. In addition, we screened 16 unique JrWD40s to walnuts based on collinearity analysis. By qRT-PCR, we found that JrWD40-133, JrWD40-150, JrWD40-155, and JrWD40-206 may regulate anthocyanin synthesis through positive regulation, whereas JrWD40-65, JrWD40-172, JrWD40-191, JrWD40-224, and JrWD40-254 may inhibit anthocyanin synthesis, suggesting that these JrWD40s are key genes affecting walnut color variation. Full article

Hyperspectral images are typically composed of hundreds of narrow and contiguous spectral bands, each containing information regarding the material composition of the imaged scene. However, these images can be affected by various sources of noise, distortions, or data loss, which can significantly degrade their quality and usefulness. This paper introduces a convergent guaranteed algorithm, LRS-PnP-DIP(1-Lip), which successfully addresses the instability issue of DHP that has been reported before. The proposed algorithm extends the successful joint low-rank and sparse model to further exploit the underlying data structures beyond the conventional and sometimes restrictive unions of subspace models. A stability analysis guarantees the convergence of the proposed algorithm under mild assumptions, which is crucial for its application in real-world scenarios. Extensive experiments demonstrate that the proposed solution consistently delivers visually and quantitatively superior inpainting results, establishing state-of-the-art performance. Full article

The Segment Anything Model (SAM) has made breakthroughs in the domain of image segmentation, attaining high-quality segmentation results using input prompts like points and bounding boxes. However, utilizing a pretrained SAM model for iris segmentation has not achieved the desired results. This is mainly due to the substantial disparity between natural images and iris images. To address this issue, we have developed SAM-Iris. First, we designed an innovative plug-and-play adapter called IrisAdapter. This adapter allows us to effectively learn features from iris images without the need to comprehensively update the model parameters while avoiding the problem of knowledge forgetting. Subsequently, to overcome the shortcomings of the pretrained Vision Transformer (ViT) encoder in capturing local detail information, we introduced a Convolutional Neural Network (CNN) branch that works in parallel with it. This design enables the model to capture fine local features of iris images. Furthermore, we adopted a Cross-Branch Attention mechanism module, which not only promotes information exchange between the ViT and CNN branches but also enables the ViT branch to integrate and utilize local information more effectively. Subsequently, we adapted SAM for iris image segmentation by incorporating a broader set of input instructions, which included bounding boxes, points, and masks. In the CASIA.v4-distance dataset, the E1, F1, mIoU, and Acc of our model are 0.34, 95.15%, 90.88%, and 96.49%; in the UBIRIS.v2 dataset, the E1, F1, mIoU, and Acc are 0.79, 94.08%, 88.94%, and 94.97%; in the MICHE dataset, E1, F1, mIoU, and Acc were 0.67, 93.62%, 88.66%, and 95.03%. In summary, this study has improved the accuracy of iris segmentation through a series of innovative methods and strategies, opening up new horizons and directions for large-model-based iris-segmentation algorithms. Full article

The demand for computer vision applications has been growing in recent years. However, most tasks in real-world scenarios are in the state of small amounts of data, and the discriminability of features results in low performance in few-shot settings. In order to obtain discriminability features in unknown classes more robustly, we first propose a simple and effective few-shot feature augmentation scheme called corner multi-head attention (CMA). The CMA module aims to project corner features through multi-head attention into a task-specific feature space through corner detection without sacrificing the generalization ability of the few-shot model. We then construct a representation space for few-shot classification by adaptively embedding the skeleton (EA). Finally, we proposed a few-shot classification framework, namely FSCA (Few-Shot Embedding Adaptation with Corner Multi-Head Attention). Its CMA module has low coupling, supports plug-and-play, and can easily be embedded into various models. To verify the effectiveness of the proposed scheme, we conducted experiments on three few-shot classification data sets, CUB-200-2011, miniImageNet, and tieredImageNet, and we performed a t-SNE clustering visual analysis. The experimental results show that it achieves continuous improvements over state-of-the-art methods on three widely used few-shot classification benchmarks. Specifically, on the miniImageNet data, the performance of FSCA under the five-way, one-shot and five-shot settings improved by an average of 1.6–4.5% and 2.2–5.4%, respectively. Full article

The Indonesian population is projected to increase by 66.65 in 2035 due to the continuous rise in urbanization globally. The growth contributed to the growing housing backlog and limited availability of residential spaces. This led to the evolution of incremental housing construction as an appropriate solution to residents’ needs. However, several factors hinder the implementation of incremental housing, including prolonged construction durations that delay the completion of an entire house, compromised quality of workmanship and materials, as well as poor flexibility. Conventional on-site construction, with concrete serving as the main material, led to prolonged construction time, difficult renovation, and untreatable waste. Preliminary studies have been conducted on incremental housing from urban development and financial perspectives, with none on alternative construction systems. Therefore, this study aimed to develop flexible and sustainable incremental housing with an assembly–disassembly system capable of reducing construction time and waste. This study experimented on the connection systems through digital simulations and prototypes leading to a construction system that combines frames and panels in a semi-volumetric system. It also combined a plug-and-play connection type to achieve the highest assembly–disassembly efficiency value (0.07), the lowest waste (below 25%), and a 30% shorter construction time. The result showed no displacement when tested with a load of up to 3 tons. This study contributed to the growing body of knowledge on alternative incremental house construction techniques, paving the way for more adaptable and environmentally responsible housing solutions in urban settings, particularly in rapidly urbanizing regions like Indonesia. Full article

Weed detection is a crucial step in achieving intelligent weeding for vegetables. Currently, research on vegetable weed detection technology is relatively limited, and existing detection methods still face challenges due to complex natural conditions, resulting in low detection accuracy and efficiency. This paper proposes the YOLOv8-EGC-Fusion (YEF) model, an enhancement based on the YOLOv8 model, to address these challenges. This model introduces plug-and-play modules: (1) The Efficient Group Convolution (EGC) module leverages convolution kernels of various sizes combined with group convolution techniques to significantly reduce computational cost. Integrating this EGC module with the C2f module creates the C2f-EGC module, strengthening the model’s capacity to grasp local contextual information. (2) The Group Context Anchor Attention (GCAA) module strengthens the model’s capacity to capture long-range contextual information, contributing to improved feature comprehension. (3) The GCAA-Fusion module effectively merges multi-scale features, addressing shallow feature loss and preserving critical information. Leveraging GCAA-Fusion and PAFPN, we developed an Adaptive Feature Fusion (AFF) feature pyramid structure that amplifies the model’s feature extraction capabilities. To ensure effective evaluation, we collected a diverse dataset of weed images from various vegetable fields. A series of comparative experiments was conducted to verify the detection effectiveness of the YEF model. The results show that the YEF model outperforms the original YOLOv8 model, Faster R-CNN, RetinaNet, TOOD, RTMDet, and YOLOv5 in detection performance. The detection metrics achieved by the YEF model are as follows: precision of 0.904, recall of 0.88, F1 score of 0.891, and mAP0.5 of 0.929. In conclusion, the YEF model demonstrates high detection accuracy for vegetable and weed identification, meeting the requirements for precise detection. Full article

In the quest to design reactors with a higher productivity, their mixing efficiency should be highly improved. The mass transfer coefficient is a parameter that measures the rate of the refining rates and has been extensively investigated in the past; however, most of the correlations developed in steelmaking are based on the effect of the gas flow rate or its alternative form, stirring energy. The gas flow rate can play a big role in mass transfer but there are many more variables involved. This work has investigated the combined effect of five variables on the mass transfer coefficient due to bottom gas injection with two injection devices: the gas flow rate, the radial position and the separation angle of the porous plugs, the slag thickness, and the ladle aspect ratio. A novel expression in a dimensionless form has been developed, which accurately predicts the mass transfer coefficient. The expression proposed indicates that increasing the gas flow rate, the slag thickness, the ladle aspect ratio, and the separation angle also increases the mass transfer coefficient. On the contrary, increasing the radial position away from the center affects mass transfer, especially at high gas flow rates. Based on the experimental data and their practical application, an optimum layout for the injection of gas is suggested to optimize both mass transfer and the mixing intensity of liquid steel. Full article

Images captured under low-light conditions often suffer from serious degradation due to insufficient light, which adversely impacts subsequent computer vision tasks. Retinex-based methods have demonstrated strong potential in low-light image enhancement. However, existing approaches often directly design prior regularization functions for either illumination or reflectance components, which may unintentionally introduce noise. To address these limitations, this paper presents an enhancement method by integrating a Plug-and-Play strategy into an extended decomposition model. The proposed model consists of three main components: an extended decomposition term, an iterative reweighting regularization function for the illumination component, and a Plug-and-Play refinement term applied to the reflectance component. The extended decomposition enables a more precise representation of image components, while the iterative reweighting mechanism allows for gentle smoothing near edges and brighter areas while applying more pronounced smoothing in darker regions. Additionally, the Plug-and-Play framework incorporates off-the-shelf image denoising filters to effectively suppress noise and preserve useful image details. Extensive experiments on several datasets confirm that the proposed method consistently outperforms existing techniques. Full article

The attention mechanism is essential to convolutional neural network (CNN) vision backbones used for sensing and imaging systems. Conventional attention modules are designed heuristically, relying heavily on empirical tuning. To tackle the challenge of designing attention mechanisms, this paper proposes a novel probabilistic attention mechanism. The key idea is to estimate the probabilistic distribution of activation maps within CNNs and construct probabilistic attention maps based on the correlation between attention weights and the estimated probabilistic distribution. The proposed approach consists of two main components: (i) the calculation of the probabilistic attention map and (ii) its integration into existing CNN architectures. In the first stage, the activation values generated at each CNN layer are modeled by using a Laplace distribution, which assigns probability values to each activation, representing its relative importance. Next, the probabilistic attention map is applied to the feature maps via element-wise multiplication and is seamlessly integrated as a plug-and-play module into existing CNN architectures. The experimental results show that the proposed probabilistic attention mechanism effectively boosts image classification accuracy performance across various CNN backbone models, outperforming both baseline and other attention mechanisms. Full article

The rise in electricity costs for households over the past year has driven significant changes in energy usage patterns, with many residents adopting smarter energy-efficient practices, such as improved indoor insulation and advanced home energy management systems powered by IoT and Digital Twin technologies. These measures not only mitigate rising bills but also ensure optimized thermal comfort and sustainability in typical residential settings. This paper proposes an innovative framework to facilitate the adoption of energy-efficient practices in households by leveraging the integration of Internet of Things technologies with Digital Twins. It introduces a novel approach that exploits standardized parametric 3D models, enabling the efficient simulation and optimization of home energy systems. This design significantly reduces deployment complexity, enhances scalability, and empowers users with real-time insights into energy consumption, indoor conditions, and actionable strategies for sustainable energy management. The results showcase that the proposed method significantly outperforms traditional approaches, achieving a 94% reduction in deployment time and a 98% decrease in memory usage through the use of standardized parametric models and plug-and-play IoT integration. Full article

With the increasing proportion of distributed generators (DGs), distribution networks usually include grid forming (GFM) and grid following (GFL) converters. However, the incompatibility of dynamic performance caused by different control methods of the GFM and GFL converters may bring instability problems and power quality risks to the distribution network. To solve this issue, the models of the GFM and GFL converters are established first to lay a good foundation for stability analysis and power quality improvement control. On this basis, an inner loop parameters design scheme is developed for GFM converters based on the D-Partition method, which facilitates the stability boundary characterization. Meanwhile, a current injection strategy is proposed to enhance the voltage support capacity of the GFL converter during grid faults. Moreover, for the distribution network with multi-converters, a compensation current control based on the analytic hierarchy process and coefficient of variation is proposed to ensure a balance between minimal capacity and optimal power quality. In this manner, DGs can be plug-and-play without considering stability and power quality issues. Finally, the effectiveness of the proposed strategy is validated with simulation results. Full article

Considerable variation in soil often occurs within and across production fields, which can significantly impact farming input management strategies. Optimizing resource utilization while enhancing crop productivity is critical for achieving Sustainable Development Goals (SDGs). This paper proposes a low-cost retrofittable Variable Rate Applicator Controller (VRAC) designed to leverage soil variability and facilitate the adoption of Variable Rate Technologies. The controller operates using a Raspberry Pi platform, RTK—Global Navigation Satellite System (GNSS), a stepper motor, and an anti-slip wheel encoder. The VRAC allows precise, on-the-fly control of the Variable Rate application of farming inputs utilizing an accurate GNSS to pinpoint geographic coordinates in real time. A wheel encoder measures accurate distance travel, providing a real-time calculation of speed with a slip-resistant wheel design for precise RPM readings. The Raspberry Pi platform processes the data, enabling dynamic adjustments of variability based on predefined maps, while the motor driver controls the motor’s RPM. It is designed to be plug-and-play, user-friendly, and accessible for a broader range of farming practices, including seeding rates, dry fertilizer, and liquid fertilizer application. Data logging is performed from various field sensors. The controller exhibits an average of 0.864 s for rate changes from 267 to 45, 45 to 241, 241 to 128, 128 to 218, and 218 to 160 kg/ha at speeds of 8, 11, 16, 19, 24, and 32 km/h. It has an average coefficient of variation of 4.59, an accuracy of 97.17%, a root means square error (RMSE) of 4.57, an R square of 0.994, and an average standard deviation of 1.76 kg for seeding discharge. The cost-effectiveness and retrofitability of this technology offer an increase in precision agriculture adoption to a broader range of farmers and promote sustainable farming practices. Full article