Search Results (2,027)

To address the limitations of the existing cotton seedling growth prediction methods in field environments, specifically, poor representation of spatiotemporal features and low visual fidelity in texture rendering, this paper proposes an algorithm for the prediction of cotton seedling growth from images based on FCA-STNet. The model leverages historical sequences of cotton seedling RGB images to generate an image of the predicted growth at time t + 1 and extracts 37 phenotypic traits from the predicted image. A novel STNet structure is designed to enhance the representation of spatiotemporal dependencies, while an Adaptive Fine-Grained Channel Attention (FCA) module is integrated to capture both global and local feature information. This attention mechanism focuses on individual cotton plants and their textural characteristics, effectively reducing the interference from common field-related challenges such as insufficient lighting, leaf fluttering, and wind disturbances. The experimental results demonstrate that the predicted images achieved an MSE of 0.0086, MAE of 0.0321, SSIM of 0.8339, and PSNR of 20.7011 on the test set, representing improvements of 2.27%, 0.31%, 4.73%, and 11.20%, respectively, over the baseline STNet. The method outperforms several mainstream spatiotemporal prediction models. Furthermore, the majority of the predicted phenotypic traits exhibited correlations with actual measurements with coefficients above 0.8, indicating high prediction accuracy. The proposed FCA-STNet model enables visually realistic prediction of cotton seedling growth in open-field conditions, offering a new perspective for research in growth prediction. Full article

Soil organic carbon (SOC) pool plays an extremely important role in regulating the global carbon (C) cycle and climate change. Atmospheric nitrogen (N) and phosphorus (P) deposition caused by human activities has significant impacts on soil C sequestration potential of terrestrial ecosystem. To investigate the effects of N and P deposition on soil C sequestration and C-N coupling relationship in broad-leaved evergreen forests, a 6-year field nutrient regulation experiment was implemented in subtropical Castanopsis sclerophylla forests with four different N and P additions: N addition (100 kg N·hm⁻²·year⁻¹), N + P (100 kg N·hm⁻²·year⁻¹ + 50 kg P·hm⁻²·year⁻¹), P addition (50 kg P·hm⁻²·year⁻¹), and CK (0 kg N·hm⁻²·year⁻¹). The changes in the C and N contents and stable isotope distributions (δ¹³C and δ¹⁵N) of different soil organic fractions were examined. The results showed that the SOC and total nitrogen (STN) (p > 0.05) increased with N addition, while SOC significantly decreased with P addition (p < 0.05), and N + P treatment has low effect on SOC, STN (p > 0.05). By density grouping, it was found that N addition significantly increased light fraction C and N (LFOC, LFN), significantly decreased the light fraction C to N ratio (LFOC/N) (p < 0.05), and increased heavy fraction C and N (HFOC, HFN) accumulation and light fraction to total organic C ratio (LFOC/SOC, p > 0.05). Contrary to N addition, P addition was detrimental to the accumulation of LFOC, LFN and reduced LFOC/SOC. It was found that different reactive oxidized carbon (ROC) increased under N addition but ROC/SOC did not change, while N + P and P treatments increased ROC/SOC, resulting in a decrease in SOC chemical stability. Stable isotope analysis showed that N addition promoted the accumulation of new soil organic matter, whereas P addition enhanced the transformation and utilization of C and N from pre-existing organic matter. Additionally, N addition indirectly increased LFOC by significantly decreasing pH; significantly contributed to LFOC and ROC by increasing STN accumulation promoted by NO₃⁻-N and NH₄⁺-N; and decreased light fraction δ¹³C by significantly increasing dissolved organic C (p < 0.05). P addition had directly significant negative effect on LFOC and SOC (p < 0.05). In conclusion, six-year N deposition enhances soil C and N sequestration while the P enrichment reduces the content of soil C, N fractions and stability in Castanopsis sclerophylla forests. The results provide a scientific basis for predicting the soil C sink function of evergreen broad-leaved forest ecosystem under the background of future climate change. Full article

The discharge of oily wastewater threatens the ecosystem and human health, and the efficient treatment of oily wastewater is confronted with problems of high mass transfer resistance at the oil-water-solid multiphase interface, significant light shielding effect, and easy deactivation of photocatalysts. Although traditional physical separation methods avoid secondary pollution by chemicals and can effectively separate floating oil and dispersed oil, they are ineffective in removing emulsified oil with small particle sizes. To address these complex challenges, photocatalytic technology and photocatalysis-based improved technologies have emerged, offering significant application prospects in degrading organic pollutants in oily wastewater as an environmentally friendly oxidation technology. In this paper, the degradation mechanism, kinetic mechanism, and limitations of conventional photocatalysis technology are briefly discussed. Subsequently, the surface interface modulation functions of metal doping and heterojunction energy band engineering, along with their applications in enhancing the light absorption range and carrier separation efficiency, are reviewed. Focus on typical studies on the separation and degradation of aqueous and oily phases using photocatalytic membrane technology, and illustrate the advantages and mechanisms of photocatalysts loaded on the membranes. Finally, other new approaches and converging technologies in the field are outlined, and the challenges and prospects for the future treatment of oily wastewater are presented. Full article

Silicon-based microcavity quantum dot lasers are attractive candidates for on-chip light sources in photonic integrated circuits due to their small size, low power consumption, and compatibility with silicon photonic platforms. However, integrating components like quantum dot lasers and photodetectors on a single chip remains challenging due to material compatibility issues and mode field mismatch problems. In this work, we have demonstrated monolithic integration of an InAs quantum dot microdisk light emitter, waveguide, and photodetector on a silicon platform using a shared epitaxial structure. The photodetector successfully monitored variations in light emitter output power, experimentally proving the feasibility of this integrated scheme. This work represents a key step toward multifunctional integrated photonic systems. Future efforts will focus on enhancing the light emitter output power, improving waveguide efficiency, and scaling up the integration density for advanced applications in optical communication. Full article

Leveraging Fourier optics theory and Abbe’s imaging principle, this study establishes that optical imaging fundamentally involves selective spatial spectrum recombination at the Fourier plane. Three classical experiments quantitatively validate universal spectrum manipulation mechanisms: (1) The Abbe-Porter experiment confirmed spectral filtering, directly demonstrating image synthesis from transmitted spectral components. (2) Zernike phase-contrast microscopy quantified spectral phase modulation, overcoming the weak-phase-object detection limit by significantly enhancing contrast. (3) Optical joint transform correlation (JTC) demonstrated efficient spectral amplitude modulation for high-speed, high-accuracy image recognition. Collectively, these results form a comprehensive framework for active light field manipulation at the spectral plane, extending modulation capabilities to phase and amplitude dimensions. This work provides a foundational theoretical and technical framework for designing advanced optical systems, extending modulation capabilities to phase and amplitude dimensions. Full article

In the rapidly evolving field of biomedical engineering, hydrogels have emerged as highly versatile biomaterials that bridge biology and technology through their high water content, exceptional biocompatibility, and tunable mechanical properties. This review provides an integrated overview of both natural and synthetic hydrogels, examining their structural properties, fabrication methods, and broad biomedical applications, including drug delivery systems, tissue engineering, wound healing, and regenerative medicine. Natural hydrogels derived from sources such as alginate, gelatin, and chitosan are highlighted for their biodegradability and biocompatibility, though often limited by poor mechanical strength and batch variability. Conversely, synthetic hydrogels offer precise control over physical and chemical characteristics via advanced polymer chemistry, enabling customization for specific biomedical functions, yet may present challenges related to bioactivity and degradability. The review also explores intelligent hydrogel systems with stimuli-responsive and bioactive functionalities, emphasizing their role in next-generation healthcare solutions. In modern medicine, temperature-, pH-, enzyme-, light-, electric field-, magnetic field-, and glucose-responsive hydrogels are among the most promising “smart materials”. Their ability to respond to biological signals makes them uniquely suited for next-generation therapeutics, from responsive drug systems to adaptive tissue scaffolds. Key challenges such as scalability, clinical translation, and regulatory approval are discussed, underscoring the need for interdisciplinary collaboration and continued innovation. Overall, this review fosters a comprehensive understanding of hydrogel technologies and their transformative potential in enhancing patient care through advanced, adaptable, and responsive biomaterial systems. Full article

In precision cotton seeding, the toothed-disk precision seeder often experiences issues with missed seeding and multiple seeding. To promptly detect and address these abnormal seeding conditions, this study develops a modular photoelectric sensing monitoring system. Initially, the monitoring time window is divided using the capacitance sensing signal between two seed drop ports. Concurrently, a photoelectric monitoring circuit is designed to convert the time when seeds block the sensor into a level signal. Subsequently, threshold segmentation is performed on the time when seeds block the photoelectric path under different seeding states. The proposed spatiotemporal joint counting algorithm identifies, in real time, the threshold type of the photoelectric sensor’s output signal within the current monitoring time window, enabling the differentiation of seeding states and the recording of data. Additionally, an STM32 micro-controller serves as the core of the signal acquisition circuit, sending collected data to the PC terminal via serial port communication. The graphical display interface, designed with LVGL (Light and Versatile Graphics Library), updates the seeding monitoring information in real time. Compared to photoelectric monitoring algorithms that detect seed pickup at the seed metering disc, the monitoring node in this study is positioned posteriorly within the seed guide chamber. Consequently, the differentiation between single seeding and multiple seeding is achieved with greater accuracy by the spatiotemporal joint counting algorithm, thereby enhancing the monitoring precision of the system. Field test results indicate that the system’s average accuracy for single-seeding monitoring is 97.30%, for missed-seeding monitoring is 96.48%, and for multiple-seeding monitoring is 96.47%. The average probability of system misjudgment is 3.25%. These outcomes suggest that the proposed modular photoelectric sensing monitoring system can meet the monitoring requirements of precision cotton seeding at various seeding speeds. Full article

The textile industry is progressively shifting towards more sustainable solutions, particularly in the field of printing technologies. This study reports the development and evaluation of water-based pigment inks formulated with bio-based pigments derived from intermediates produced via bacterial fermentation. Two pigments—indigo (blue) and quinacridone (red)—were incorporated into ink formulations and applied on cotton and polyester fabrics through valve-jet inkjet printing (ChromoJet). The physical properties of the inks were analyzed to ensure compatibility with the equipment, and printed fabrics were assessed as to their color fastness to washing, rubbing, artificial weathering, and artificial light. The results highlight the good performance of the bio-based inks, with excellent light and weathering fastness and satisfactory wash and rub resistance. The effect of different pre-treatments, including a biopolymer and a synthetic binder, was also investigated. Notably, the biopolymer pre-treatment enhanced pigment fixation on cotton, while the synthetic binder improved wash fastness on polyester. These findings support the integration of biotechnologically sourced pigments into eco-friendly textile digital printing workflows. Full article

HPMC, regulating slurry properties, is widely used in cement-based materials. Research on the application of HPMC in gangue slurry is still in its early stages. Moreover, the interactive effects of various factors on gangue slurry performance have not been thoroughly investigated. The work examined the effects of slurry concentration (X₁), maximum gangue particle size (X₂), and HPMC dosage (X₃) on slurry performance using response surface methodology (RSM). The microstructure of the slurry was characterized via scanning electron microscopy (SEM) and polarized light microscopy (PLM), while low-field nuclear magnetic resonance (LF-NMR) was employed to analyze water distribution. Additionally, industrial field tests were conducted. The results are presented below. (1) X₁ and X₃ exhibited a negative correlation with layering degree and slump flow, while X₂ showed a positive correlation. Slurry concentration had the greatest impact on slurry performance, followed by maximum particle size and HPMC dosage. HPMC significantly improved slurry stability, imposing the minimum negative influence on fluidity. Interaction terms X₁X₂ and X₁X₃ significantly affected layering degree and slump flow, while X₂X₃ significantly affected layering degree instead of slump flow. (2) Derived from the RSM, the statistical models for layering degree and slump flow define the optimal slurry mix proportions. The gangue gradation index ranged from 0.40 to 0.428, with different gradations requiring specific slurry concentration and HPMC dosages. (3) HPMC promoted the formation of a 3D floc network structure of fine particles through adsorption-bridging effects. The spatial supporting effect of the floc network inhibited the sedimentation of coarse particles, which enhanced the stability of the slurry. Meanwhile, HPMC only converted a small amount of free water into floc water, which had a minimal impact on fluidity. HPMC addition achieved the synergistic optimization of slurry stability and fluidity. (4) Field industrial trials confirmed that HPMC-optimized gangue slurry demonstrated significant improvements in both stability and flowability. The optimized slurry achieved blockage-free pipeline transportation, with a maximum spreading radius exceeding 60 m in the goaf and a maximum single-borehole backfilling volume of 2200 m³. Full article

Owing to the intricate variability of underwater environments, images suffer from degradation including light absorption, scattering, and color distortion. However, U-Net architectures severely limit global context utilization due to fixed-receptive-field convolutions, while traditional attention mechanisms incur quadratic complexity and fail to efficiently fuse spatial–frequency features. Unlike local enhancement-focused methods, HMENet integrates a transformer sub-network for long-range dependency modeling and dual-domain attention for bidirectional spatial–frequency fusion. This design increases the receptive field while maintaining linear complexity. On UIEB and EUVP datasets, HMENet achieves PSNR/SSIM of 25.96/0.946 and 27.92/0.927, surpassing HCLR-Net by 0.97 dB/1.88 dB, respectively. Full article

Rice lodging severely affects crop growth, yield, and mechanized harvesting efficiency. The accurate detection and quantification of lodging areas are crucial for precision agriculture and timely field management. However, Unmanned Aerial Vehicle (UAV)-based lodging detection faces challenges such as complex backgrounds, variable lighting, and irregular lodging patterns. To address these issues, this study proposes SWRD–YOLO, a lightweight instance segmentation model that enhances feature extraction and fusion using advanced convolution and attention mechanisms. The model employs an optimized loss function to improve localization accuracy, achieving precise lodging area segmentation. Additionally, a grid-based lodging ratio estimation method is introduced, dividing images into fixed-size grids to calculate local lodging proportions and aggregate them for robust overall severity assessment. Evaluated on a self-built rice lodging dataset, the model achieves 94.8% precision, 88.2% recall, 93.3% mAP@0.5, and 91.4% F1 score, with real-time inference at 16.15 FPS on an embedded NVIDIA Jetson Orin NX device. Compared to the baseline YOLOv8n-seg, precision, recall, mAP@0.5, and F1 score improved by 8.2%, 16.5%, 12.8%, and 12.8%, respectively. These results confirm the model’s effectiveness and potential for deployment in intelligent crop monitoring and sustainable agriculture. Full article

The accurate extraction of navigational regions from images of navigational waters plays a key role in ensuring on-water safety and the automation of unmanned vessels. Nonetheless, current technological methods encounter significant challenges in addressing fluctuations in water surface illumination, reflective disturbances, and surface undulations, among other disruptions, in turn making it challenging to achieve rapid and precise boundary segmentation. To cope with these challenges, in this paper, we propose a coordinate-aware multi-scale feature network (GASF-ResNet) method for water segmentation. The method integrates the attention module Global Grouping Coordinate Attention (GGCA) in the four downsampling branches of ResNet-50, thus enhancing the model’s ability to capture target features and improving the feature representation. To expand the model’s receptive field and boost its capability in extracting features of multi-scale targets, the Avoidance Spatial Pyramid Pooling (ASPP) technique is used. Combined with multi-scale feature fusion, this effectively enhances the expression of semantic information at different scales and improves the segmentation accuracy of the model in complex water environments. The experimental results show that the average pixel accuracy (mPA) and average intersection and union ratio (mIoU) of the proposed method on the self-made dataset and on the USVInaland unmanned ship dataset are 99.31% and 98.61%, and 98.55% and 99.27%, respectively, significantly better results than those obtained for the existing mainstream models. These results are helpful in overcoming the background interference caused by water surface reflection and uneven lighting in the aquatic environment and in realizing the accurate segmentation of the water area for the safe navigation of unmanned vessels, which is of great value for the stable operation of unmanned vessels in complex environments. Full article

Traditional approaches in tunnel lighting have been directed toward the installation of appropriate luminaires in the intermediate and transitional sections with the simple objective of diminishing the effect of delayed visual accommodation during daylight hours. Such efforts run in parallel with the target of keeping the huge electrical use at the lowest level. Nevertheless, inadequate attention has been conceded to the interior areas, whose noticeable longitude in several instances, and subsequently the duration of occupancy of the users, can produce discomfort in the majority of the tunnel or underground passageway. It is in this region where the flicker effect presents a more remarkable impact. Although such effect is in fact uncomfortable, the strategies to eliminate it efficiently have not been developed in depth and the result is still deserving, especially in terms of sustainability. The reasons for this neglect, as well as some particularities and solutions, are exposed and discussed in the present article. Specifically, it is proved that the use of sunlight can be an adequate initiative and a positive energy input into design and retrofit tunnels capable of hampering or totally avoiding such unwanted effect. The innovative tunnel geometry explained in this manuscript is not cylindrical, and it is not based in revolution forms. Thus, it prevents the appearance of such unnerving visual effects, which compromise sustainability and endanger security. We are in the position to explain how the vector field generated by the normal to the points of the novel surface displayed remains non-parallel, ensuring appropriate diffusivity and, consequently, an even distribution of radiated energy. In the same manner, the notion of the tunnel is extended from a linear system to a veritable network of galleries, which can traverse in space bi- or even three-dimensionally. Accordingly, we will offer diverse instances of junctions and splices that further enhance the permeability into the terrain, augmenting the resilience capabilities of this disruptive technology. With all the former, a net reduction of costs reaching 25% can be easily expected with revenues. Full article

To address the challenges of low recognition accuracy and substantial model complexity in crop disease identification models operating in complex field environments, this study proposed a novel hybrid model named ConvTransNet-S, which integrates Convolutional Neural Networks (CNNs) and transformers for crop disease identification tasks. Unlike existing hybrid approaches, ConvTransNet-S uniquely introduces three key innovations: First, a Local Perception Unit (LPU) and Lightweight Multi-Head Self-Attention (LMHSA) modules were introduced to synergistically enhance the extraction of fine-grained plant disease details and model global dependency relationships, respectively. Second, an Inverted Residual Feed-Forward Network (IRFFN) was employed to optimize the feature propagation path, thereby enhancing the model’s robustness against interferences such as lighting variations and leaf occlusions. This novel combination of a LPU, LMHSA, and an IRFFN achieves a dynamic equilibrium between local texture perception and global context modeling—effectively resolving the trade-offs inherent in standalone CNNs or transformers. Finally, through a phased architecture design, efficient fusion of multi-scale disease features is achieved, which enhances feature discriminability while reducing model complexity. The experimental results indicated that ConvTransNet-S achieved a recognition accuracy of 98.85% on the PlantVillage public dataset. This model operates with only 25.14 million parameters, a computational load of 3.762 GFLOPs, and an inference time of 7.56 ms. Testing on a self-built in-field complex scene dataset comprising 10,441 images revealed that ConvTransNet-S achieved an accuracy of 88.53%, which represents improvements of 14.22%, 2.75%, and 0.34% over EfficientNetV2, Vision Transformer, and Swin Transformer, respectively. Furthermore, the ConvTransNet-S model achieved up to 14.22% higher disease recognition accuracy under complex background conditions while reducing the parameter count by 46.8%. This confirms that its unique multi-scale feature mechanism can effectively distinguish disease from background features, providing a novel technical approach for disease diagnosis in complex agricultural scenarios and demonstrating significant application value for intelligent agricultural management. Full article

Ghost imaging (GI) facilitates image acquisition under low-light conditions through single pixel measurements, thus holding tremendous potential across various fields such as biomedical imaging, remote sensing, defense and military applications, and 3D imaging. However, in order to reconstruct high-resolution images, GI typically requires a large number of single-pixel measurements, which imposes practical limitations on its application. Parallel ghost imaging addresses this issue by utilizing each pixel of a position-sensitive detector as a bucket detector to simultaneously perform tens of thousands of ghost imaging measurements in parallel. In this work, we explore the non-local characteristics of ghost imaging in depth, and by constructing a large speckle space, we achieve a reconstruction result in parallel ghost imaging where the field of view surpasses the limitations of the reference arm detector. Using a computational ghost imaging framework, after pre-recording the speckle patterns, we are able to complete X-ray ghost imaging at a speed of 6 min per sample, with image dimensions of 14,000 × 10,000 pixels (4.55 mm × 3.25 mm, millimeter-scale field of view) and a pixel resolution of 0.325 µm (sub-micron pixel resolution). We present this framework to enhance efficiency, extend resolution, and dramatically expand the field of view, with the aim of providing a solution for the practical implementation of ghost imaging. Full article