Search Results (582)

Crop targets in UAV aerial images are typically characterized by small scale, dense distribution, severe mutual occlusion, and complex backgrounds, which often lead to low detection accuracy and large counting errors for existing deep learning models. To address these issues, this study proposes an improved YOLOv12-based crop detection and counting model, named YOLO-DC. By introducing an attention mechanism (LGCB-AM) and a multi-scale detection head (MS-DH), the proposed model effectively enhances local texture extraction, global modeling, foreground–background contrast, and boundary perception for dense small objects. Subsequently, a series of comparative experiments, ablation studies, and transfer experiments were conducted on the wheat and rice datasets. The results show that YOLO-DC achieves a favorable balance among detection accuracy, counting error, and model efficiency and overall outperforms the other comparison models. Ablation studies further verify the effectiveness of the proposed design, showing that LGCB-AM is the key contributor to the performance improvement, while the boundary branch and repulsion branch play critical roles in dense-target discrimination. In addition, an appropriate module insertion strategy can effectively balance high-level semantic enhancement and feature fusion stability. Transfer experiments demonstrate that pretraining on the wheat dataset and fine-tuning on the rice dataset significantly outperform training from scratch, indicating strong cross-crop transfer potential. Overall, the proposed YOLO-DC provides an effective solution for high-precision crop detection and counting in agricultural scenarios. Full article

Extracting high-molecular-weight (HMW) DNA from cactus tissues remains technically challenging due to the abundance of mucilage, pectins, polyphenols, and other metabolites that compromise DNA purity, increase viscosity, and reduce integrity, thereby limiting its suitability for long-read sequencing. This constraint is particularly relevant in Selenicereus megalanthus, a crop of increasing agronomic and genomic importance for which optimized protocols for third-generation sequencing remain limited. Here, we compared four CTAB-based DNA extraction protocols using dehydrated cladode tissue and evaluated DNA quality using NanoDrop spectrophotometry, Qubit fluorometry, agarose gel electrophoresis, and functional validation via sequencing on the Oxford Nanopore PromethION 2 Solo platform. Among the tested methods, our proposed optimized mucilage-adapted CTAB (MuCTAB) protocol, comprising 4% CTAB, 4% PVP-40, 0.5% β-mercaptoethanol, and proteinase K, showed the best overall performance. MuCTAB yielded the highest dsDNA concentration (239.63 ± 34.37 ng/µL), optimal purity ratios (A260/A280 = 1.96 ± 0.05; A260/A230 = 2.01 ± 0.01), and superior DNA integrity. Nanopore validation confirmed its effectiveness, with MuCTAB producing the highest sequencing yield (84.2 Gbp), read count, mean read quality score (Q18.4), read N50 (40.3 kbp), and maximum read length (1.9 Mbp). Overall, MuCTAB represents a low-cost, reproducible, and efficient method for HMW DNA extraction from mucilage-rich pitahaya tissues and shows promising potential for adaptation to other recalcitrant plant species. Full article

Root phenotypic analysis is closely related to crop yield and stress resistance. Although deep learning can improve the efficiency of root phenotype recognition, existing methods suffer from insufficient segmentation accuracy under complex soil backgrounds and focus on a single target. To address the issues of limited accuracy and operational complexity in existing root segmentation models, this paper proposes a novel wavelet-enhanced full-scale segmentation network. The WaveUNet+ model is based on U-Net3plus, replaces traditional downsampling with the Haar wavelet transform, and introduces the EMA module. The impact of the wavelet transform is validated using Grad-CAM, and HD95 is employed to evaluate the improvement in segmentation quality brought by the attention mechanism from the perspective of boundary accuracy. Transfer learning is used to improve model generalization, and the test results on diverse roots and various soils are compared. A Docker containerized root image segmentation method is designed to achieve convenient and practical operation, and the deployment feasibility of the model on edge devices is also verified. Our model effectively enhances the recognition of fine roots in soil backgrounds, leading to improvements across various metrics, achieving an Accuracy of 99.2%, while improving model accuracy with relatively low parameter count and model size. Compared with the original U-Net model, mIoU is increased by 1.52% and Recall by 2.93%. The results show that the model not only performs excellently on the original dataset but also maintains good generalization ability across different imaging modalities, crop species, and soil conditions. With Docker, users can achieve root image segmentation on their own computers without tedious program installation and environment configuration. In the future, we will attempt methods such as pruning and quantization to reduce model size, so as to better adapt to the deployment requirements of edge devices. Full article

This paper presents Geo-Esthetics, an application-oriented workflow that uses remote sensing imagery as source morphology for generative design. The study addresses a design problem: how can large-scale terrestrial textures be extracted, abstracted, and organized as pattern references for parametric and visual design? Nine representative geomorphological settings were selected. For each case, Sentinel-2 imagery was cropped into a 2 km × 2 km geographic window, enhanced using spectral-index selection and Contrast Limited Adaptive Histogram Equalization (CLAHE), and used as an image prompt in Midjourney v6.0. A consistent prompt structure and parameter setting were applied. Four variants were generated for each case and screened according to topological fidelity, level of abstraction, and design applicability. Box-counting dimension and lacunarity were calculated to compare morphological complexity between source images and generated patterns. The cases show that hydrological, tectonic, desert, agricultural, and reef morphologies can be translated into design-oriented pattern prototypes for paving, façades, interfaces, acoustic elements, and biomimetic surfaces. The contribution of this work lies mainly in design methodology: it provides a documented workflow for connecting Earth observation data, generative AI, and design ideation, while retaining clear boundaries around model reproducibility, prompt sensitivity, case representativeness, and perceptual evaluation. Full article

LeafScans-Orchard is a curated, multi-year RGB image dataset of orchard plant leaves designed to support research in computer vision, machine learning, and plant phenotyping. The dataset comprises 9708 high-quality leaf scans acquired during collection campaigns conducted between 2015 and 2025, covering seven orchard crop species: apple, pear, sweet cherry, sour cherry, plum, peach, and apricot. In total, the dataset includes 67 cultivar labels. All samples were acquired using flatbed scanning under controlled conditions on a uniform background, ensuring high visual consistency and minimal background variability. The original scans were captured at 1200 dpi and subsequently converted into a public release format at 300 dpi, stored as lossless TIFF images to preserve morphological and textural details. Each image corresponds to a single leaf and is organized in a hierarchical directory structure by species, cultivar, and acquisition year, accompanied by image-level metadata and aggregated species–cultivar–year counts. LeafScans-Orchard is suitable for plant species classification, cultivar recognition, leaf morphology analysis, texture analysis, and general visual feature extraction. In addition to the main release, a representative subset of 300 original 1200 dpi scans is provided to support high-resolution analyses. The dataset is particularly suited for fine-grained classification, morphology-driven analysis, and methodological studies under controlled imaging conditions. Full article

Mycotoxin contamination of cereal crops poses escalating food safety risks across the Central and Balkan European (CBE) corridor under climate change, yet no PRISMA 2020-compliant synthesis of remote sensing (RS) and machine learning (ML) evidence for this region exists. We conducted an AI-augmented systematic review applying a four-stage automated pipeline—PICO domain scoring, SBERT semantic deduplication, and Thompson-sampling reinforcement learning—to 36,038 corpus records (2010–2025), yielding 156 included studies (inter-rater κ = 0.81 (95% CI: 0.74–0.88)). Logistic growth modelling identified a 56-fold corpus expansion with inflection at t₀ = 2024.8 (R² = 0.981). Satellite multispectral imaging dominated the literature (91.7% of studies); random forest and gradient boosting models achieved R² = 0.74–0.80 for aflatoxin B1 and deoxynivalenol prediction in CBE maize and wheat when integrating vegetation indices, land surface temperature, and precipitation covariates. Deep learning surpassed classical ML in annual study count from 2021, reaching ~60% relative share by 2025, though the performance advantage narrows at field scale relative to laboratory hyperspectral benchmarks (98–99% accuracy). A five-percentage-point CBE–global performance gap is largely consistent with differences in sample size and multi-toxin design scope rather than algorithmic access. The country × mycotoxin gap matrix identifies zero eligible studies for four CBE nations and for T-2/HT-2 toxins across the Balkan states. Climate-driven satellite mycotoxin prediction emerges as the field’s active research frontier. Full article

For modern agricultural management, the accuracy of plant identification is crucial. However, the task becomes challenging because crops and weeds at early growth stages often exhibit similar color, leaf morphology, and texture in two-dimensional images captured under field conditions, despite their clear biological differences in terms of botanical species, root systems, and phenological characteristics. Furthermore, computing hardware in the field also has strict limits. Therefore, we developed the RT-BMTR network to handle these physical constraints. Within this architecture, image data is processed through a bilateral hybrid backbone named Bi-HMB. The DSFM captures small local details, and MambaVision understands the broader background information. Then, these features are fused by RepNCSPELAN4. We adopted this structure to reduce redundant calculations. Next, the model determines its bounding boxes using the Inner-ShapeIoU loss function. This geometric constraint improves the detection of small targets. When evaluated on the CropAndWeed dataset, our model achieved an average precision (AP) at IoU threshold 0.5 (AP⁵⁰) of 68.1%, AP⁷⁵ of 54.8%, and a mean AP averaged over IoU thresholds from 0.5 to 0.95 (AP^50–95) of 50.9%. Detection precision recorded 26.5% for small objects, 44.7% for moderate ones, and with 59.3% for large objects. Rates for the first two categories saw enhancements of 16.2% and 4.6%. Overall, our modified model outperforms the original RT-DETR baseline. We also shrank the overall parameter count by 30.1%, alongside a 4.2% decrease in computational demand. Full article

Anaerobic co-digestion (AcoD) is increasingly applied in sewage-sludge management and organic-waste treatment because it can improve methane recovery, stabilize mixed substrates, and reduce life-cycle greenhouse-gas emissions under appropriate feedstock and operating conditions. However, existing reviews still focus mainly on feedstock types or isolated enhancement measures and less often connect synergistic mechanisms with engineering implementation and carbon outcomes. The specific novelty of this review is to connect functional feedstock classification, mechanism boundaries, engineering controls, and carbon-performance evaluation within one sludge-centered AcoD framework. This review synthesizes recent progress in AcoD of sewage sludge, food waste, livestock manure, crop residues, and industrial organic streams through a chain from feedstock traits to substrate interactions, microbial responses, engineering performance, and carbon benefits. Feedstocks are reorganized by function rather than by waste name, highlighting how carbon-to-nitrogen contrast, buffering capacity, hydrolysis recalcitrance, and inhibitor profiles jointly define synergy potential. Key mechanisms, including C/N balancing, hydrolysis complementarity, inhibitor mitigation, and direct interspecies electron transfer (DIET), are discussed together with their applicability limits. Representative evidence shows methane-yield or methane-production increases of about 41–55% for selected food-waste–manure blends, approximately 45% for rice–straw–pig manure systems after cellulolytic pretreatment, and approximately 16–55% for selected additive strategies; these values are illustrative rather than directly comparable because the underlying studies differ in substrates, baselines, reactor configurations, pretreatment conditions, and operating parameters. The review then translates mechanism into practice through pretreatment, reactor-selection templates, operating windows, additive reinforcement, and artificial-intelligence-assisted monitoring. Representative cases and life-cycle evidence indicate that AcoD can improve methane productivity while lowering greenhouse-gas emissions relative to landfill or mono-digestion pathways when energy substitution and nutrient recycling are credibly counted. Remaining bottlenecks include incomplete kinetic integration, limited DIET quantification, insufficient reporting of quantitative operating ranges and additive dosages, and weak coupling of carbon, economics, and regional feedstock dynamics. The revised review therefore treats AcoD as a sludge-centered mechanism-to-engineering framework and highlights two transferability gaps that require stronger standardization: biodegradation/toxicity testing and local co-substrate logistics. Full article

Duckweed (Lemnaceae) is a promising alternative feed crop, particularly in regions with nutrient surpluses and protein deficits, as it grows efficiently on nutrient-rich agricultural wastewater and provides protein-rich biomass. However, its high moisture content and rapid post-harvest spoilage pose major storage challenges. This study evaluated (co-)ensiling as a cost-effective preservation strategy for duckweed. Three separate experiments were conducted to assess the ensilability of duckweed alone and in combination with various agricultural co-substrates and additives, including corn silage, beet pulp, grass silage, hemp shives, hay, molasses, sun-dried duckweed and CaCO₃. Duckweed alone could not be successfully ensiled due to excessive moisture, resulting in poor acidification and high levels of undesirable fermentation products. During the long-term co-ensiling test, a duckweed–corn silage mixture containing 29% fresh duckweed and 71% corn silage showed the most stable fermentation profile, with low pH, limited fermentation losses, and no detectable butyric acid. A duckweed–grass silage mixture containing 51% fresh duckweed and 49% grass silage allowed higher duckweed inclusion and retained the highest level of apparent pepsin-digestible protein after storage, but showed elevated acetic acid and ethanol concentrations. A duckweed–beet pulp mixture containing 74% fresh duckweed and 26% beet pulp enabled the highest duckweed inclusion rate, but showed signs of clostridial fermentation, likely due to excess moisture. Microbiological analysis of this beet pulp mixture showed reduced Enterobacteriaceae after ensiling, but also increased clostridial counts. Oxalic acid concentrations were low in all duckweed-based silages, with the largest reduction observed in the duckweed–grass mixture. Overall, the results show that duckweed co-ensiling is feasible but highly dependent on co-substrate selection and moisture control. Further formulation optimisation is required, particularly for high-duckweed mixtures, to reduce the risk of clostridial fermentation and improve practical applicability as a storable feed ingredient. Full article

Currently, devices such as solar insecticidal lamps are widely used in agricultural pest control, but routine trapping is insufficient to meet the demands of precision agriculture. Therefore, determining the nocturnal phototactic rhythm of pests to optimize the control strategies of insecticidal lamps has become key to achieving precise pest control. However, existing automated monitoring and forecasting devices struggle to effectively monitor the nocturnal phototactic rhythm of small pests. To address this issue, this study developed an automated monitoring system for phototactic rhythm based on sticky traps and machine vision. For the hardware, an image acquisition device integrating a darkroom and scheduled supplementary lighting was designed to obtain stable time-series images of nocturnal pests. For the algorithm, the YOLO-STP detection model was proposed by improving upon the baseline YOLOv11 model. This model introduces a P2 detection layer, a Coordinate Attention (CA) mechanism, and a hybrid bounding box regression loss function integrating WIoU and NWD. Combined with a sliding window cropping method, it further enhances the detection capability for small objects. Additionally, an incremental counting method based on spatial cascade matching was proposed to mitigate counting errors caused by target occlusion or detachment in the time-series images. Experimental results indicate that the mean average precision (mAP) of the detection model was 93.2%. For the counting method, the coefficient of determination ($R^2$) was 0.98, with an RMSE of 1.97 and an MAE of 1.60. Field validation in real-world paddy fields demonstrated that the system can accurately record the abundance changes of 12 pest species, intuitively visualizing the differences in phototactic rhythms among various species. This study provides a viable automated monitoring tool for acquiring the nocturnal activity rhythm data of agricultural pests in the field. Full article

Persistent traffic congestion and the need for efficient traffic monitoring have increased the demand for automated vehicle-analysis systems based on CCTV footage. This study presents a CCTV-based vehicle monitoring system that integrates vehicle detection, tracking, counting, public/private vehicle class prediction, seven-category vehicle-type prediction, vehicle-color recognition, and traffic-state estimation using YOLOv12 and DeepSORT. To reduce manual annotation effort during the initial training stage, a semi-automated method for generating synthetic composite road scenes was developed by combining cropped vehicle images and road-background images. The detector was first trained on 10,000 synthetic images and then sequentially fine-tuned on real CCTV data. Four real-world traffic video clips from Metro Manila were used in the study. Three 5 min clips were used within the staged refinement workflow: the first two for iterative refinement and the third for final post-refinement evaluation of the adapted model. A separate fourth CCTV clip was reserved exclusively for blind evaluation without on-the-fly retraining. The final system achieved average accuracies of 97% for public/private vehicle class prediction, 90% for seven-category vehicle-type prediction, 82% for vehicle-color recognition, and 96.67% for vehicle counting on the final evaluation video. The results show that synthetic pretraining combined with limited real-world fine-tuning can improve performance in CCTV-based vehicle monitoring while reducing the amount of manually labeled real-world data required. The study also discusses the limitations of the current evaluation protocol and the need for broader multi-location testing. Full article

Growth chambers are vital in controlled environment agriculture (CEA), enabling precise regulation of environmental conditions for year-round crop production, especially in urban areas with limited arable land. This study retrofitted a nonfunctional cold room into a plant growth chamber with controlled temperature, humidity, and CO₂ levels to evaluate lettuce (Lactuca sativa) growth under three LED treatments: broad-spectrum white, combined white and far-red, and narrow amber (598 nm). Environmental parameters were controlled at 21 °C during the day and 19 °C at night, with 65% relative humidity, and 800 ppm CO₂. After 40 days, plants under combined white and far-red LEDs produced the tallest shoots (21.8 ± 0.3 cm) and highest leaf count (23.7 ± 0.9). No significant differences were observed among treatments for fresh mass, dry mass, head diameter, or relative chlorophyll content. The findings demonstrated the feasibility of retrofitting a nonfunctional cold room into a controlled environment growth chamber capable of supporting lettuce cultivation under the tested conditions. Full article

Zeugodacus tau, a highly destructive agricultural quarantine pest causing severe economic losses to global fruit crops, urgently requires the development of male fertility-based control strategies. Here, we identified and characterized the testis-specific serine/threonine protein kinase 1 gene (ZtTSSK1) in Z. tau. The encoded protein of ZtTSSK1 is highly conserved among dipteran species. Spatiotemporal expression analysis revealed predominant expression in adult males, with specific localization to the testes. In situ hybridization further localized ZtTSSK1 transcripts to the transformation zone. Furthermore, functional characterization by RNA interference (RNAi) revealed that knockdown of ZtTSSK1 resulted in a significant 62% reduction in sperm counts. While egg numbers laid by females mated to dsZtTSSK1- versus dsGFP-injected males did not differ, hatching rates were significantly lower for eggs from dsZtTSSK1 matings. These findings establish ZtTSSK1 as a key regulator of spermatogenesis and male fertility in Z. tau, providing a theoretical foundation and candidate target for genetic-based sterile insect technique (gSIT) development. Full article

Agricultural crop residue burning is a major driver of seasonal PM2.5 pollution and greenhouse gas (GHG) emissions in Northern Thailand. This study quantified GHG emissions from the open burning of rice, maize, and sugarcane residues across six provinces (Chiang Mai, Mae Hong Son, Lampang, Uttaradit, Nakhon Sawan, and Kamphaeng Phet) from 2019 to 2024 using the 2006 IPCC emission methodology. Spatiotemporal patterns of fire hotspots were characterized using MODIS and VIIRS satellite data, combined with kernel density estimation (KDE) and land-use classification in ArcGIS Pro. Total non-CO₂ GHG emissions (CH₄ and N₂O, expressed as CO₂-eq using GWP100 from IPCC AR5) over the six years totaled 2,599,551 tCO₂-eq, with major rice contributing the largest share (35%), followed by sugarcane (24%), second rice (21%), and maize (20%). Nakhon Sawan was the leading emitter (41%), reflecting its extensive rice and sugarcane cultivation. Pearson correlation analysis revealed consistently positive relationships between daily fire hotspot counts and PM2.5 concentrations (r = 0.30–0.84), with the strongest correlations observed in Mae Hong Son, where basin topography traps pollutants. Time-series analysis confirmed pronounced seasonal PM2.5 peaks that exceeded Thailand’s 24-h NAAQS limit (37.5 μg/m³) by 7–9 times in severe years. Biochar production via pyrolysis was evaluated as a zero-burning alternative, with an estimated annual carbon sequestration potential of 2.3–3.5 million tCO₂-eq, substantially exceeding emissions from open burning. These findings indicate that crop-residue valorization options—including biochar production, composting, and biochar co-compost—could theoretically offset agricultural GHG emissions and reduce field-burning PM2.5 emissions in Northern Thailand. However, the realized mitigation will depend on (i) verification of biochar long-term stability in tropical Thai soils through dedicated in situ trials, (ii) economic incentives that offset biochar production costs of approximately 1500–3500 THB per tonne, and (iii) integration within a policy mix that combines burning bans, mechanization support, and farmer extension services. Without these enabling conditions, biochar should be regarded as a future-perspective option rather than an immediately deployable solution. Full article

Early frost during the R6 and R7 maturity stages of soybean (Glycine max L.) usually causes immature (green or semi-green) crops to be harvested. These immature soybean seeds have a shrunken appearance, green tone, and high chlorophyll content in the oil, leading to heavy discounts for farmers at the elevator. Previous lab-scale storage studies have shown that seed color can change under light and warm temperatures; however, light cannot be added to a commercial storage bin. Therefore, this study examined the effect of field drying and storage conditions on immature soybean color and oil quality. Soybean planted in two plots were desiccated at the R6 and R7 maturity stages and then allowed to field dry. The field-dried desiccated soybeans were conditioned to moisture contents (MCs) of 12 and 17% and stored in airtight plastic bags at respective temperatures of 4 °C and 22.5 °C for 24 weeks. Seed color, mold, and oil quality were analyzed at intervals of 0, 4, 8, 16, and 24 weeks. The desiccated R6 seeds’ color “a” value significantly changed during field drying from (−9.75 to +0.19) and (−8.96 to +1.95) for Plot 1 and Plot 2, respectively. This means that the color changed from green to a golden yellow or light greenish-brown color after field drying. The chlorophyll content of the desiccated soybeans after field drying at the two maturity stages for both plots was less than 3 mg kg⁻¹ of oil and was relatively stable throughout storage. During storage, at 17% moisture content and 22.5 °C, mold counts increased significantly for R6, R7, and R8 (frozen) control soybeans between weeks 0 and 4 to 4.36 CFU g⁻¹, 5.93 CFU g⁻¹ and 6.22 CFU g⁻¹, respectively. Peroxide and free fatty acid values were within acceptable limits across all storage temperatures and moisture contents. This study suggests that favorable weather conditions for field drying after an early frost have the potential to improve the color of harvested and stored soybeans, similar to mature soybeans. Full article