
CA-GAN generates realistic synthetic weed images for site-specific weed management, achieving 82.63% and 93.46% recognition accuracy on public and Japanese datasets. #WeedManagement #DeepLearning #AgTech Details: doi.org/10.34133/plantphenom…

Plant Phenomics, indexed in: #DOAJ #EI #PMC #SCIE (#JIF 2026: 8.2) #Scopus (#CiteScore2026: 10.6) etc. #PlantPhenomics #PlantPhenotyping #openaccess

CA-GAN generates realistic synthetic weed images for site-specific weed management, achieving 82.63% and 93.46% recognition accuracy on public and Japanese datasets. #WeedManagement #DeepLearning #AgTech Details: doi.org/10.34133/plantphenom…

RIFSeg-Net fuses RGB and infrared UAV imagery to segment soybean canopies, enabling automated classification of leaf types and dynamic phenotyping for breeding. #SoybeanPhenotyping #UAV #AgTech Details: doi.org/10.34133/plantphenom…

Polarization-based algorithm decouples leaf surface glare from tissue scattering, improving vegetation index accuracy by an order of magnitude for field phenotyping. #PlantPhenotyping #RemoteSensing #AgTech Details: doi.org/10.34133/plantphenom…

PAT pipeline automates periderm length quantification in Arabidopsis roots using deep learning, achieving 94% efficiency and enabling large-scale genetic screens. #PlantDevelopment #HighThroughputPhenotyping #Arabidopsis Details: doi.org/10.34133/plantphenom…

FRCNN-A with global attention improves wheat spike detection in dense canopies, achieving 85% mAP on augmented data while remaining lightweight. #WheatPhenotyping #DeepLearning #AgTech Details: doi.org/10.34133/plantphenom…

Semi-supervised learning with resampling enables accurate rubber leaf nitrogen and potassium classification from hyperspectral data, even with small imbalanced datasets. #RubberTree #Hyperspectral #MachineLearning Details: doi.org/10.34133/plantphenom…

Systematic review of 262 studies reveals deep learning dominance in plant stress detection, driven by open-source imaging repositories like PlantVillage and Kaggle. #PlantStress #DeepLearning #AI Details: doi.org/10.34133/plantphenom…

Frequency-domain analysis of chlorophyll fluorescence dynamics reveals NPQ effects below 0.03 rad/s, enabling compensation for measurement variability during strong actinic illumination. #ChlorophyllFluorescence #PlantPhysiology #Photosynthesis Details: doi.org/10.34133/plantphenom…

Improved CycleGAN with spatial-coordinate-based background separation expands in situ root datasets, boosting Improved_UNet generalization by 33.6% mIoU for root segmentation. #RootPhenotyping #DeepLearning #AgTech Details: doi.org/10.34133/plantphenom…

Multi-target regression chaining improves hyperspectral prediction of 10 tomato leaf element concentrations, boosting R² for Mn, Cu, Co, Fe, and Mg. #Hyperspectral #MachineLearning #PlantNutrition Details: doi.org/10.34133/plantphenom…

ALS-based 3D detection quantifies understory saplings beneath closed canopies, enabling precision forest gap management and improved light-resource utilization for regeneration. #ForestRegeneration #LiDAR #Forestry Details: doi.org/10.34133/plantphenom…

Combining total canopy PRI with SIF via two-leaf modeling improves rice GPP estimation (R²=0.88), outperforming hotspot observations for tracking crop productivity. #SIF #GPP #RemoteSensing Details: doi.org/10.34133/plantphenom…

Hyperspectral imaging with machine learning enables nondestructive rapeseed maturity classification, achieving 97.86% accuracy using D2nd-IVISSA-SPA-SVM model. #OilseedRape #Hyperspectral #agtech Details: doi.org/10.34133/plantphenom…

DomAda-FruitDet enables cross-species auto-labeling via domain-adaptive anchor-free detection, achieving 88-94% average precision for apples, tomatoes, pitaya, and mangoes. #FruitDetection #AutoLabeling #agtech Details: doi.org/10.34133/plantphenom…

Unified framework for structural and practical identifiability analysis in mechanistic plant growth models, introducing new risk index based on profile likelihood confidence intervals. #PlantModeling #ParameterEstimation #SystemsBiology Details: doi.org/10.34133/plantphenom…

CNN models estimate root length from minirhizotron images without segmentation, achieving R² up to 0.986 across four crop species and stress conditions. #RootPhenotyping #DeepLearning #PrecisionAg Details: doi.org/10.34133/plantphenom…