Search Results (217)

UAV small target detection in urban security, disaster monitoring, agricultural inspection, and other fields faces the challenge of increasing accuracy and real-time requirements. However, existing detection algorithms still have weak small target representation ability, extensive computational resource overhead, and poor deployment adaptability. Therefore, this paper proposes a lightweight algorithm, I-YOLOv11n, based on YOLOv11n, which is systematically improved in terms of both feature enhancement and structure compression. The RFCBAMConv module that combines deformable convolution and channel–spatial attention is designed to adjust the receptive field and strengthen the edge features dynamically. The multiscale pyramid of STCMSP context and the lightweight Transformer–DyHead hybrid detection head are designed by combining the multiscale hole feature pyramid (DFPC), which realizes the cross-scale semantic modeling and adaptive focusing of the target area. A collaborative lightweight strategy is proposed. Firstly, the semantic discrimination ability of the teacher model for small targets is transferred to guide and protect the subsequent compression process by integrating the mixed knowledge distillation of response alignment, feature imitation, and structure maintenance. Secondly, the LAMP–Taylor channel pruning mechanism is used to compress the model redundancy, mainly to protect the key channels sensitive to shallow small targets. Finally, K-means++ anchor frame optimization based on IoU distance is implemented to adapt the feature structure retained after pruning and the scale distribution of small targets of UAV. While significantly reducing the model size (parameter 3.87 M, calculation 14.7 GFLOPs), the detection accuracy of small targets is effectively maintained and improved. Experiments on VisDrone, AI-TOD, and SODA-A datasets show that the mAP@0.5 and mAP@0.5:0.95 of I-YOLOv11n are 7.1% and 4.9% higher than the benchmark model YOLOv11 n, respectively, while maintaining real-time processing capabilities, verifying its comprehensive advantages in accuracy, light weight, and deployment. Full article

The present-day issues related to the cotton-growing industry, namely yield estimation, pest effect, and growth phase diagnostics, call for integrated, scalable monitoring solutions. This write-up reveals Cotton Multitask Learning (CMTL), a transformer-driven multitask framework that launches three major agronomic tasks from UAV pictures at one go: boll detection, pest damage segmentation, and phenological stage classification. CMTL does not change separate pipelines, but rather merges these goals using a Cross-Level Multi-Granular Encoder (CLMGE) and a Multitask Self-Distilled Attention Fusion (MSDAF) module that both allow mutual learning across tasks and still keep their specific features. The biologically guided Stage Consistency Loss is the part of the architecture of the network that enables the system to carry out growth stage transitions that occur in reality. We executed CMTL on a tri-source UAV dataset that fused over 2100 labeled images from public and private collections, representing a variety of crop stages and conditions. The model showed its virtues state-of-the-art baselines in all the tasks: setting 0.913 mAP for boll detection, 0.832 IoU for pest segmentation, and 0.936 accuracy for growth stage classification. Additionally, it runs at the fastest speed of performance on edge devices such as NVIDIA Jetson Xavier NX (Manufactured in Shanghai, China), which makes it ideal for deployment. These outcomes evoke CMTL’s promise as a single and productive instrument of aerial crop intelligence in precision cotton agriculture. Full article

The utilisation of agrifood waste ashes has the potential to enhance the nutrient content of cereal crops, thereby optimising both yield and grain quality. This study investigated rye grain composition, the fermentation efficiency, and volatile compounds in mashes made from crops fertilised with agrifood waste ashes derived from the combustion of corn cob, wood chips, and biomass with defecation lime. The ashes were applied at 2, 4, and 8 t/ha, separately and as mixtures of corn cob (25%) with wood chips (75%) and corn cob (50%) with biomass and defecation lime (50%). Rye mashes were prepared using the pressureless starch liberation method. The starch content in the majority of the rye grains was comparable to the control sample (57.12 g/100 g). The range of ethanol concentrations observed in the fermented mashes was from 55.55 to 68.12 g/L, which corresponded to fermentation yields of 67.25–76.59% of theoretical. The lowest fermentation yield was exhibited by the mash derived from rye cultivated on soil fertilised with a 50:50 mixture of ashes from corn cob and biomass with defecation lime at 8 t/ha. This mash contained more than double the acetaldehyde concentration and total aldehyde content compared to the other samples. These findings demonstrate the potential of using waste biomass ash as a source of macro- and microelements for rye cultivation, enabling the production of agricultural distillates. To ensure high fermentation efficiency and low aldehyde levels, ash dosage and composition need to be established based on experimental optimisation. Full article

In recent years, great attention has been paid to second-generation (from agricultural and industrial wastes) lactic acid (LA) production. In the present study, the possibility of two Lactiplantibacillus plantarum strains, namely 53 and 2HS, to produce LA from waste materials was investigated. Distiller’s dried grains with solubles (DDGS), spent coffee grounds (SCG), wood chips, and cheese whey were used as substrates after pretreatment, and the results were compared with those with lactose as a carbon source. Both strains were capable of assimilating sugars from all waste materials. Nearly 20 g/L LA from 23 g/L reducing sugars (RS) obtained from DDGS, 22 g/L LA from 21 g/L RS from SCG, and 22 g/L LA from 21 g/L whey lactose were produced compared to 22 g/L LA obtained from 22 g/L lactose monohydrate in the fermentation broth. The wood chip hydrolysate (WH) contains only 10 g/L RS, and its fermentation resulted in the production of 5 g/L LA. This amount is twice as low as that produced from 11 g/L lactose monohydrate. A mathematical model was constructed based on the Compertz and Luedeking–Piret equations. Full article

The feasibility of utilizing biochar as a static floating material for agricultural applications was researched to prevent evaporation from open water static storage systems or as a floating barrier in slurry pits, for instance. Five types of biochar were created from chips, bark, and pellets of pine and residues from two acacia species using a pyrolysis time between 60 and 120 min and mean temperatures between 380 and 690 °C in a simple double-chamber reactor. Biomass and biochar were characterized for their main properties: bulk density, moisture content, volatile matter, ash content, fixed carbon, and pH. Biochar was also evaluated through a basic floatability test over 27 days (648 h) in distilled water. The highest fixed carbon content was observed in pine bark biochar (69.5%), followed by the pine pellets (67.4%) and pine chips (63.4%). Despite their high carbon content, the pellets exhibited a low floatability level, whereas pine bark biochar showed superior static floatage times, together with chip and ground chip biochar. These results suggest that biochar produced from bark and wood chips may be suitable for application as floatability material in water or slurry management systems. These results warrant further research into the static floating of biochar. Full article

The combined effect of non-thermal plasma (NTP) and hydropriming on the germination performance and seedling characteristics of specific varieties of cucumber (Cucumis sativus L.) and broccoli (Brassica oleracea var. italica Plenck.) seeds after short-term storage is reported. Seeds were treated with NTP for 10 and 15 min, followed by hydropriming in distilled water for 24 h, and then stored for six months in the dark before evaluation. The treated cucumber seeds demonstrated a statistically significant enhancement in seed germination and seedling vitality indices. In contrast, broccoli seeds showed no significant improvement. The stimulatory effects observed in cucumber may be attributed to reactive oxygen and nitrogen species, which act as signaling molecules to promote stress tolerance and early growth. This study also highlights the potential of combined NTP treatment and hydropriming as a pre-sowing treatment for select crops, underscoring the need for species-specific optimization. The used, portable, and relatively inexpensive NTP device offers practical advantages for agricultural applications. Full article

Recent progress in geospatial foundation models (GFMs) has demonstrated strong generalization capabilities for remote sensing downstream tasks. However, existing GFMs still struggle with fine-grained cropland classification due to ambiguous field boundaries, insufficient and low-efficient temporal modeling, and limited cross-regional adaptability. In this paper, we propose CropSTS, a remote sensing foundation model designed with a decoupled temporal–spatial attention architecture, specifically tailored for the temporal dynamics of cropland remote sensing data. To efficiently pre-train the model under limited labeled data, we employ a hybrid framework combining joint-embedding predictive architecture with knowledge distillation from web-scale foundation models. Despite being trained on a small dataset and using a compact model, CropSTS achieves state-of-the-art performance on the PASTIS-R benchmark in terms of mIoU and F1-score. Our results validate that structural optimization for temporal encoding and cross-modal knowledge transfer constitute effective strategies for advancing GFM design in agricultural remote sensing. Full article

Plant diseases significantly undermine agricultural productivity. This study introduces an improved YOLOv10n model named WD-YOLO (Weighted and Double-scale YOLO), an advanced architecture for efficient plant disease detection. The PlantDoc dataset was initially enhanced using data augmentation techniques. Subsequently, we developed the DSConv module—a novel convolutional structure employing double-scale weighted convolutions that dynamically adjust to different scale perceptions and optimize attention allocation. This module replaces the conventional Conv module in YOLOv10. Furthermore, the WTConcat module was introduced, dynamically merging weighted concatenation with a channel attention mechanism to replace the Concat module in YOLOv10. The training of WD-YOLO incorporated knowledge distillation techniques using YOLOv10l as a teacher model to refine and compress the architectural learning. Empirical results reveal that WD-YOLO achieved an mAP50 of 65.4%, outperforming YOLOv10n by 9.1% without data augmentation and YOLOv10l by 2.3%, despite having significantly fewer parameters (9.3 times less than YOLOv10l), demonstrating substantial gains in detection efficiency and model compactness. Full article

This review examines the chemical functionalization of Camelina, hemp, and rapeseed oils for the development of sustainable bio-based resins. Key strategies, including epoxidation, acrylation, and click chemistry, are discussed in the context of tailoring molecular structure to enhance reactivity, compatibility, and material performance. Particular emphasis is placed on overcoming the inherent limitations of vegetable oil structures to enable their integration into high-performance polymer systems. The agricultural sustainability and environmental advantages of these feedstocks are also highlighted alongside the technical challenges associated with their chemical modification. Functionalized oils derived from Camelina, hemp, and rapeseed have been successfully applied in various resin systems, including protective coatings, pressure-sensitive adhesives, UV-curable oligomers, and polyurethane foams. These advances demonstrate their growing potential as renewable alternatives to petroleum-based polymers and underline the critical role of structure–property relationships in designing next-generation sustainable materials. Ultimately, the objective of this review is to distill the most effective functionalization pathways and design principles, thereby illustrating how Camelina, hemp, and rapeseed oils could serve as viable substitutes for petrochemical resins in future industrial applications. Full article

Maize (Zea mays L.), a vital global food crop, relies on its stomatal structure for regulating photosynthesis and responding to drought. Conventional manual stomatal detection methods are inefficient, subjective, and inadequate for high-throughput plant phenotyping research. To address this, we curated a dataset of over 1500 maize leaf epidermal stomata images and developed a novel lightweight detection model, StomaYOLO, tailored for small stomatal targets and subtle features in microscopic images. Leveraging the YOLOv11 framework, StomaYOLO integrates the Small Object Detection layer P2, the dynamic convolution module, and exploits large-scale epidermal cell features to enhance stomatal recognition through auxiliary training. Our model achieved a remarkable 91.8% mean average precision (mAP) and 98.5% precision, surpassing numerous mainstream detection models while maintaining computational efficiency. Ablation and comparative analyses demonstrated that the Small Object Detection layer, dynamic convolutional module, multi-task training, and knowledge distillation strategies substantially enhanced detection performance. Integrating all four strategies yielded a nearly 9% mAP improvement over the baseline model, with computational complexity under 8.4 GFLOPS. Our findings underscore the superior detection capabilities of StomaYOLO compared to existing methods, offering a cost-effective solution that is suitable for practical implementation. This study presents a valuable tool for maize stomatal phenotyping, supporting crop breeding and smart agriculture advancements. Full article

In this paper, we propose a knowledge distillation framework specifically designed for semantic segmentation tasks in agricultural scenarios. This framework aims to address several prevalent challenges in smart agriculture, including limited computational resources, strict real-time constraints, and suboptimal segmentation accuracy on cropped images. Traditional single-level feature distillation methods often suffer from insufficient knowledge transfer and inefficient utilization of multi-scale features, which significantly limits their ability to accurately segment complex crop structures in dynamic field environments. To overcome these issues, we propose a multi-level distillation strategy that leverages feature and embedding patch distillation, combining high-level semantic features with low-level texture details for joint distillation. This approach enables the precise capture of fine-grained agricultural elements, such as crop boundaries, stems, petioles, and weed clusters, which are critical for achieving robust segmentation. Additionally, we integrated an enhanced attention mechanism into the framework, which effectively strengthens and fuses key crop-related features during the distillation process, thereby further improving the model’s performance and image understanding capabilities. Extensive experiments on two agricultural datasets (sweet pepper and sugar) demonstrate that our method improves segmentation accuracy by 7.59% and 6.79%, without significantly increasing model complexity. Further validation shows that our approach exhibits strong generalization capabilities on two widely used public datasets, proving its applicability beyond agricultural domains. Full article

Accurate detection of Jinxiu Malus fruits in unstructured orchard environments is hampered by frequent overlap, occlusion, and variable illumination. To address these challenges, we propose YOLOv8-MSP-PD (YOLOv8 with Multi-Scale Pyramid Fusion and Proportional Distance IoU), a lightweight model built on an enhanced YOLOv8 architecture. We replace the backbone with MobileNetV4, incorporating unified inverted bottleneck (UIB) modules and depth-wise separable convolutions for efficient feature extraction. We introduce a spatial pyramid pooling fast cross-stage partial connections (SPPFCSPC) module for multi-scale feature fusion and a modified proportional distance IoU (MPD-IoU) loss to optimize bounding-box regression. Finally, layer-adaptive magnitude pruning (LAMP) combined with knowledge distillation compresses the model while retaining performance. On our custom Jinxiu Malus dataset, YOLOv8-MSP-PD achieves a mean average precision (mAP) of 92.2% (1.6% gain over baseline), reduces floating-point operations (FLOPs) by 59.9%, and shrinks to 2.2 MB. Five-fold cross-validation confirms stability, and comparisons with Faster R-CNN and SSD demonstrate superior accuracy and efficiency. This work offers a practical vision solution for agricultural robots and guidance for lightweight detection in precision agriculture. Full article

Agricultural practices significantly influence agroecosystem biodiversity, driving a growing focus on the development of environmentally sustainable management strategies. Olive (Olea europaea L.) is one of the most widely cultivated tree crops in the Mediterranean basin and other regions with a Mediterranean climate. In this study, we employed a split-plot design with whole plots arranged as a randomized complete block design (RCBD) to evaluate the effects of minimum tillage and the application of wood distillate to olive canopies on wild vascular plant and soil-dwelling springtail communities in a conventionally managed olive grove in central Italy. Biotic communities were sampled twice, in November and April. Tillage caused a marginally significant decrease in springtail species richness in April and significantly influenced the composition of both plant and springtail communities in April. All the plant species showed a decrease in abundance under tillage, whereas the abundance of springtail species responded to tillage in a species-specific way. Wood distillate had no effect on any community attribute in either season. Springtail total abundance was not affected by any treatment in either season. Our findings confirm that tillage practices affect the diversity of plant and springtail communities. Moreover, we had evidence that spring tillage may have more negative impacts on the studied communities with respect to autumn tillage. Moreover, we suggest that the application of low-concentration wood distillate to olive canopies can be considered, in the short-term, a sustainable agricultural practice that does not negatively affect agroecosystem biodiversity. Full article

Tagetes erecta L. is an ornamental crop known for its medicinal qualities. Large amounts of waste are produced in the commercial usage of T. erecta flowers, including leaves that could be used to develop new eco-friendly phenolic extracts with additional value for the food industry. To maximize the phenol content in the leaf extracts, this study used a Box–Behnken design with Response Surface Methodology, considering three extraction methods (Soxhlet distillation, heat, and vacuum-assisted extraction), three cropping practices (without fertilizer, chemical fertilizer, and vermicompost), and three phenological stages (plants without buds, with buds, and in flower). Extracts from plants fertilized with vermicompost (Eisenia foetida, 10 t ha⁻¹), collected during the blossoming stage and extracted via Soxhlet distillation, exhibited the highest phenol content (25.66 mg GAE/g). Further chemical characterization of the optimized extract (UV-Vis, UV-fluorescence, FTIR, GC-MS, HPLC) confirmed the occurrence of polyphenols in the extract, including quercetin, chlorogenic, gallic, p-coumaric, 3-hydroxycinnamic, and caffeic acids. This underscores the significance of T. erecta leaf residues as a valuable source of bioactive molecules, highlighting the importance of integrating agricultural practices and chemical extraction methods to enhance the phenolic content in leaf extracts from this species. Full article

This study explores the production of anhydrous ethanol from discarded fruits, aiming to determine optimal fermentation conditions and evaluate the feasibility of a green separation technology. Fermentation experiments were performed using juices from Psidium guajava (S1), Carica paapaya (S2), and mucilage residues of Coffea arabica (S3). All fermentations were carried out at a pH of 4.5 for 7 days in 1 L bioreactors. A full 2² factorial design was applied to evaluate the effects of two variables: yeast type (commercial Saccharomyces cerevisiae [CY] vs. native yeast [NY]) and temperature (21 °C vs. 30 °C). Higher ethanol concentrations were achieved with CY at 30 °C, yielding 6.79% ethanol for S3. A multi-criteria matrix prioritized coffee residues due to their high ethanol yield, biomass availability, and economic viability. The ethanol was dehydrated using a packed-bed bioadsorption system with crushed corn, which increased purity from 6.7% v/v to 98.9% v/v in two stages, while avoiding azeotropic limitations. Energy analysis revealed low specific consumption (3.68 MJ/kg), outperforming conventional distillation. The results of this study, obtained at operating temperatures of 30 °C and 21 °C, a pH of 4.5, and an operating time of 7 days in a 1L bioreactor, demonstrate ethanol concentrations of 6.79%, confirming the technical feasibility of using agricultural waste as a raw material and validating the efficiency of a bioadsorption-based dehydration system. These findings address the current gap in integrating green ethanol separation with low-cost agricultural residues and highlight a sustainable alternative for decentralized bioethanol production. Full article