Field-based phenotyping of water-related traits for precision irrigation in tropical agroecosystems poses a persistent methodological challenge, driven by high climatic variability and the complex water-use physiology of Crassulacean Acid Metabolism (CAM) crops such as pineapple (Ananas comosus var. MD2). We developed and validated a Physics-Informed Machine Learning (PIML) framework that integrates high-resolution UAV multispectral imagery, IoT-based microclimatic records, and a mechanistic soil water balance based on the FAO-56 Penman–Monteith standard to predict plot-scale soil moisture depletion as a proxy of plant water status. A six-month field campaign (March–August 2022) across 25 georeferenced commercial pineapple plots in the Colombian Orinoquia piedmont yielded a spatiotemporally balanced dataset of N=150 observations. Soil-adjusted vegetation indices (OSAVI, MSAVI) outperformed standard NDVI for capturing water-related canopy traits, effectively decoupling spectral responses from substrate noise. A Gradient Boosting regressor achieved R2=0.842 and RMSE=0.0705 on a normalized target scale, corresponding to a 7.05% error over the prediction range, while the traffic-light Decision Support System (DSS) for irrigation scheduling reached 91.1% accuracy (Cohen’s Kappa =0.91). Incorporating daily soil moisture depletion as a mechanistic feature improved predictive accuracy over a spectral-only baseline (ΔR2=+0.052) and anchored predictions within a physically consistent framework based on the FAO-56 water balance, with no false negatives observed for water deficit detection in the hold-out validation set. This framework advances high-throughput, population-scale phenotyping of water-related traits in open-canopy CAM crops, establishing a transferable methodology for operational precision irrigation under tropical savanna conditions.
Why it matches plant phenotyping methodsUAVマルチスペクトル画像と機械学習を用いて植物の水関連形質・水状態を推定する枠組みを開発・検証しており、表現型取得と予測手法が研究の中心である。
abstractWe developed and validated a Physics-Informed Machine Learning (PIML) framework that integrates high-resolution UAV multispectral imagery, IoT-based microclimatic records, and a mechanistic soil water balance based on the FAO-56 Penman–Monteith standard to predict plot-scale soil moisture depletion as a proxy of plant water status.
Reproduction assets foundThe paper's Data Availability Statement explicitly deposits the complete dataset and source code (raw UAV multispectral imagery, Python scripts, IoT sensor logs, CROPWAT 8.0 files, and XGBoost model code) in a public Mendeley Data repository, which directly reproduces this paper's phenotyping measurements and analysis.Dataset · publicThe complete dataset and source code supporting this study are publicly available at Mendeley Data: https://data.mendeley.com/datasets/9xwdvzf3bf/1 (accessed on 20 May 2026). The repository includes: (1) raw multispectral UAV imagery with calibration panel captures; (2) Python scripts for DN-to-reflectance conversion and spectral index extraction; (3) IoT sensor logs (soil moisture, temperature, relative humidity); (4) CROPWAT 8.0 project files for FAO-56 soil water balance simulation; and (5) XGBoost model source code with hyperparameter optimization routines.Open asset ↗Mendeley Data · 9xwdvzf3bf/1lines:193-228Plant phenotyping relevance match · UnverifiedEurope PMC · checked 5 Sept 2026
Phytophthora cinnamomi is considered as one of the world's worst plant pathogens, infecting about 5,000 plant species including those of agricultural and environmental significance. Disease management is largely dependent on chemical control, particularly synthetic fungicides such as phosphonic acid-based fungicides, e.g., phosphite/potassium phosphonate. While phosphonic-acid-based fungicides have been highly effective for more than 40 years, their prolonged use has led to the development of tolerance and decreased sensitivity in P. cinnamomi . Novel control agents that are effective but environmentally sustainable are therefore urgently needed. RNA-based biopesticides, which use exogenously applied double-stranded RNA (dsRNA) specific to the target pest or pathogen to avoid off-target effects on other organisms in the environment including beneficials, have emerged as a potential novel disease management strategy against P. cinnamomi . Due to the limited availability of bioassays to study the efficacy of this novel control agent against P. cinnamomi , we developed water-based lupin and pineapple bioassays using readily available plastic cups and glassware with mycelial plugs as inoculum. Infection rate was assessed 3 to 7 days post-inoculation (dpi) for lupin and 7 to 14 dpi for pineapple by measuring root lesion length and rating root rot. Potassium phosphonate (Agri Fos 600) and dsRNA were tested as example control agents, with dsRNA uptake tested via northern blotting. The bioassays were found suitable for P. cinnamomi pathogenicity assays, with one mycelial plug an effective inoculum; fungicide sensitivity testing, with doses as low as 0.45 g L -1 Agri Fos® 600 providing protection; and exogenous dsRNA studies targeting root pathogens, with dsRNA able to be taken up by germinating lupin seeds. Overall, the assays are soil-free and thus overcome dsRNA stability issues in the soil and enable the collection of intact clean roots for molecular analyses. Furthermore, the bioassays are non-destructive, allowing root lesion symptoms to be visually monitored and repeatedly measured across different timepoints.
Why it matches plant phenotyping methods植物病害の根病徴を測定する土壌フリー・バイオアッセイを開発し、その適用性を検証しており、表現型取得法が研究の中心です。
abstractwe developed water-based lupin and pineapple bioassays using readily available plastic cups and glassware with mycelial plugs as inoculum.
Near-infrared (NIR) imaging plays a crucial role in precision agriculture; however, the high cost of multispectral sensors limits its widespread adoption. In this study, we generate synthetic NIR images (2592 × 1944 pixels) of pineapple crops from standard RGB drone imagery using the Pix2PixHD framework. The model was trained for 580 epochs, saving the first model after epoch 1 and then every 10 epochs thereafter. While models trained beyond epoch 460 achieved marginally higher metrics, they introduced visible artifacts. Model 410 was identified as the most effective, offering consistent quantitative performance while producing artifact-free results. Evaluation of Model 410 across 229 test images showed a mean SSIM of 0.6873, PSNR of 29.92, RMSE of 8.146, and PCC of 0.6565, indicating moderate to high structural similarity and reliable spectral accuracy of the synthetic NIR data. The proposed approach demonstrates that reliable NIR information can be obtained without expensive multispectral equipment, reducing costs and enhancing accessibility for farmers. By enabling advanced tasks such as vegetation segmentation and crop health monitoring, this work highlights the potential of deep learning–based image translation to support sustainable and data-driven agricultural practices. Future directions include extending the method to other crops, environmental conditions and real-time drone monitoring.
Why it matches plant phenotyping methodsRGB画像から作物モニタリング用の合成NIR画像を生成する画像変換手法を開発し、テスト画像で定量評価しているため、植物情報取得法が中心である。
abstractwe generate synthetic NIR images (2592 × 1944 pixels) of pineapple crops from standard RGB drone imagery using the Pix2PixHD framework.
Pineapple leaf fibers (PALFs) are sustainable resources with exceptional tenacity, yet their component-structure-property relationships remain underexplored, limiting high-value applications. This study establishes quantitative links between chemical composition (cellulose, hemicellulose, lignin) and mechanical properties of single PALFs, aiming to develop a predictive model for rapid fineness and strength assessment. Using stepwise chemical degumming, we generated 11 distinct fiber groups (SSD1–11) from Queen PALF and characterized > 600 fibers via standardized mechanical testing (GB/T5881–2024). Pearson and Mantel correlation analysis revealed a hierarchical component-function framework: cellulose governs PALF stiffness via crystalline microfibrils; hemicellulose modulates fineness and interfacial adhesion as bonding network; lignin enhances stretchability and strength via stress-transfer structure. Critically, we developed a computational model, termed Prediction of Fineness and Strength of Single PALF (PFS-PALF) for rapid assessment, which was experimentally validated to achieve 95 % similarity versus national standard measurements. This approach has potential to replace the conventional labor-intensive and tedious measurements on PALF’s mechanical behavior. In addition, PFS-PALF enables reverse regulation of PALFs’ composition through optimized degumming parameters and facilitates rapid selection of suitable PALF variety to meet application-specific mechanical requirements.
Why it matches plant phenotyping methodsパイナップル葉繊維の細さ・引張強度を迅速推定する計算モデルを開発し、標準測定との一致度で実験検証しており、表現型測定手法が研究の中心である。
abstractwe developed a computational model, termed Prediction of Fineness and Strength of Single PALF (PFS-PALF) for rapid assessment, which was experimentally validated to achieve 95 % similarity versus national standard measurements.
Food quality detection by machine learning (ML) is more practical and sustainable as it does not require sample preparation and reagents. However, the prediction of pineapple quality by hyperspectral data applied to ML is not known. The aim of this study was to verify accurate ML models for predicting pineapple fruit quality and the best inputs for algorithms: Artificial Neural Networks (ANNs), M5P (model tree), REPTree decision trees, Random Forest (RF), Support Vector Machine (SMV) and Zero R. Three inputs were used for each model: leaf reflectance, peel reflectance, and fruit reflectance. The machine learning model SVM, stood out for its best results, demonstrating good generalization capacity and effectiveness in predicting these attributes, reaching accuracy values above 0.7 for Brix and ratio, using fruit reflectance. In terms of the overall efficiency of the input variables, peel and fruit were the most informative, with peel standing out for the estimation of secondary metabolism compounds, while the fruit showed excellent performance in predicting flavor-related attributes, such as acidity, °Brix and ratio, as mentioned previously, above 0.7. These results highlight the potential of using spectral data and machine learning in the non-destructive assessment of pineapple quality, enabling advances in monitoring and selecting fruits with better sensors.
Why it matches plant phenotyping methodsハイパースペクトル反射データと機械学習によるパイナップル果実品質形質の非破壊推定を検証しており、形質取得・推定法が研究の中心である。
abstractThe aim of this study was to verify accurate ML models for predicting pineapple fruit quality and the best inputs for algorithms
Spectral vegetation indices are used for non-invasive monitoring of plant health and evaluation of plant nitrogen status, chlorophyll content, and green leaf biomass. In particular, normalized difference vegetative index (NDVI) is commonly used to evaluate plant health by measuring reflected near infrared (NIR) light against visible red light. This information is applied to improve the precision of fertilizer application, irrigation, and other field treatment activities. However, commercially available spectral sensors are cost-intensive, which limits its accessibility to small-scale farmers and further spectral vegetation index studies in developing countries. Thus, we developed a portable, low cost, open source spectral sensor, and validated its accuracy against a more established but expensive reference spectral sensor (SKR 1840(ND) with data logger component CR1000). Our low cost spectral sensor consists of SparkFun RedBoard microcontroller that is programmable in Arduino IDE. The device is mounted on a 1.8 pole to ensure similar sensor height from the reference sensor used. The program was developed and uploaded using open source Arduino Software (IDE) version 1.8.15 Additional components include Sparkfun Qwiic OpenLog (compatible from 64 MB to 32 GB microSD card) and SparkFun Real Time Clock Module (RV-1805 Qwiic). For the spectral sensors, both upward and downward facing SparkFun Spectral Sensor Breakout AS7263 NIR and AS7262 VIS spectral sensors were used. Here, we present the field validation of our portable low cost spectral sensor and the result of the year-long NDVI measurement of pineapple plants under various pineapple residue treatments in our Philippine field site. Plant height, number of leaves and leaf chlorophyll concentration were also measured in parallel to compare NDVI measurements from various pineapple residue treatments. Similar low cost sensor was developed and deployed for NDVI measurements in the field site in Southern Benin to evaluate rice crop development under various N fertilizer amount and water management practices.
Why it matches plant phenotyping methods低コストNDVIセンサーを開発し、既存センサーとの精度検証および圃場での植物状態測定に用いており、センサーによる表現型取得が中心である。
abstractThus, we developed a portable, low cost, open source spectral sensor, and validated its accuracy against a more established but expensive reference spectral sensor
PineappleField / plotFruitClassificationObject detectionGrowth / development / phenology
Pineapples are an important agricultural economic crop in Taiwan. Considerable human resources are required to protect pineapples from excessive solar radiation, which could otherwise lead to overheating and subsequent deterioration. Note that simple covering all of the fruit with a paper bag is not a viable solution, due to the fact that it makes it impossible to determine whether the fruit is ripe. This paper proposes a system by which to automate the detection of ripe pineapples. The proposed deep learning architecture enables detection regardless of lighting conditions, achieving accuracy of more than 99.27% with error of less than 2% at distances of 300 ~ 800 mm. This proposed system using an Nvidia TX2 is capable of 15 frames per second, thereby making it possible to mount the device on machines that move at walking speed.
Why it matches plant phenotyping methodsパイナップル果実の成熟状態を画像から推定する深層学習システムの開発と性能評価が中心であり、成熟という植物器官の状態を直接推定するため、植物フェノタイピング手法に該当する。
abstractThis paper proposes a system by which to automate the detection of ripe pineapples.
Natural flowering affects fruit development and quality, and impacts the harvest of specialty plants like pineapple. Pineapple growers use chemicals to induce flowering so that most plants within a field produce fruit of high quality that is ready to harvest at the same time. Since pineapple is hand-harvested, the ability to harvest all of the fruit of a field in a single pass is critical to reduce field losses, costs, and waste, and to maximize efficiency. Traditionally, due to high planting densities, pineapple growers have been limited to gathering crop intelligence through manual inspection around the edges of the field, giving them only a limited view of their crop's status. Through the advances in remote sensing and computer vision, we can enable the regular inspection of the field and automated inflorescence counting enabling growers to optimize their management practices. Our work uses a deep learning-based density estimation approach to count the number of flowering pineapple plants in a field with a test MAE of 11.5 and MAPD of 6.37%. Notably, the computational complexity of this method does not depend on the number of plants present and therefore efficiently scale to easily detect over a 1.6 million flowering plants in a field. We further embed this approach in an active learning framework for continual learning and model improvement.
Why it matches plant phenotyping methodsパイナップルの開花状態を画像から自動計数・位置推定する深層学習手法が研究の中心であり、精度評価と継続学習も行っているため。
abstractOur work uses a deep learning-based density estimation approach to count the number of flowering pineapple plants in a field with a test MAE of 11.5 and MAPD of 6.37%.
Abstract Pineapple is an economically important tropical fruit crop, but the lack of adequate planting material limits its productivity. A range of micropropagation protocols has been developed over the years to address this shortfall. Still, the final stage of micropropagation, i.e. acclimatisation, remains a challenge as pineapple plantlets grow very slowly. Several studies have been conducted focusing on this phase and attempting to improve plantlet growth and establishment, which requires tools for the non-destructive evaluation of growth during acclimatisation. This report describes the use of semi-automated and automated image analysis to quantify canopy growth of pineapple plantlets, during five months of acclimatisation. The canopy area progressively increased during acclimatisation, particularly after 90 days. Regression analyses were performed to determine the relationships between the automated image analysis and morphological indicators of growth. The mathematical relationships between estimations of the canopy area and the fresh and dry weights of intact plantlets, middle-aged leaves (D leaves) and roots showed determination coefficients (R2) between 0.84 and 0.92. We propose an appropriate tool for the simple, objective and non-destructive evaluation of pineapple plantlets growth, which can be generally applied for plant phenotyping, to reduce costs and develop streamlined pipelines for the assessment of plant growth.
Why it matches plant phenotyping methodsパイナップル苗のキャノピー画像から成長を非破壊・自動推定する方法を開発・検証しており、植物表現型取得が研究の中心です。
abstractThis report describes the use of semi-automated and automated image analysis to quantify canopy growth of pineapple plantlets