Search

• Data-driven tools for environmental regulation models in protected agriculture. • Independent software platform for processing Pn data and formulating targets. • Functions for Pn prediction, environment regulation model development, and visualization. • Ability to handle photosynthetic data, achieving high precision and low error. Protected agriculture has emerged as a key solution to address the pressing issue of food scarcity. To enhance crop yield, environmental regulation techniques have been widely employed in protected production. However, the absence of user-friendly, data-driven tools for developing regulation models remains a challenge. This study aims to propose IPECM, an independent and user-friendly software platform for processing and analyzing crop photosynthetic rate (Pn) data and formulating environmental regulation targets. The platform provides functionalities, such as Pn prediction model development, environmental regulation model development and result visualization, supporting various machine learning algorithms and regulation target obtaining algorithms. The IPECM Platform's application is demonstrated through examples of light intensity regulation for cucumber growth and CO 2 concentration regulation for tomato growth. The results showcase the software's ability to handle photosynthetic data of any dimension, with the established Pn prediction model achieving a coefficient of determination of 0.98 and a root mean square error lower than 1 μmol·m−2·s−1. The established regulation models can achieve maximum Pn or optimal energy utilization efficiency according to user requirements. IPECM Platform is an independent, automated, and open-source software for protected environmental regulation modeling, providing both the modeling process and results visualization. It offers valuable services for protected agriculture research, eliminating the need for programming knowledge. [ABSTRACT FROM AUTHOR]

• A portable device was developed to detect moldy cores in apples. • Available spectral shape features are selected through one-way analysis of variance. • A new model was constructed that fused spectral characteristics and shapes. • The device can detect samples with moldy-core degrees less than 6%. Moldy core is a disease that significantly affects apple yield. However, discriminating slightly moldy cores is a substantial challenge in actual production. In this study, a device for detecting moldy cores in apples is developed. The device is supported by a C12880MA sensor and an STM32F103 microcontroller to detect Vis/NIR signals of apples. Experimentally obtained spectral data of Fuji apples were collected to analyze the characteristic wavelengths through different preprocessing methods for the detection of moldy cores. Additionally, multiple spectral shape features were extracted according to peaks and troughs. Significant differences in spectral shape features between healthy and moldy core samples were analyzed via a one-way analysis of variance. Finally, an apple moldy core discrimination model was constructed using an adaptive boosting algorithm that fused spectral characteristics and shapes. The model accuracy of the training set and the test set was 99.1 % and 97.3 %, respectively. Compared with other models, the proposed model effectively improved the accuracy of detecting samples with moldy core degrees less than 6 %. This research provides a novel method for detecting moldy cores, which is of great significance for apple quality management. [ABSTRACT FROM AUTHOR]

• Proposing a framework for multi-plant phenotype extraction with minimal point clouds. • Non-parametric strategy enhances point cloud spatial representation capability. • Creating a labeled cotton point cloud dataset with multi-temporal and multi-species. Rapid and precise 3D organ segmentation is crucial for the automatic extraction of phenotypic traits, forming a fundamental prerequisite for intelligent plant breeding. The advancement of deep learning technology has replaced labor-intensive manual measurements and traditional computer vision methods, which are sensitive to parameters, in phenotypic trait extraction. However, current larger network structures not only require extensive point cloud data but also consume substantial computational resources, rendering them unsuitable for agricultural tasks with limited plant samples. Therefore, this study developed a lightweight 3D deep learning network (PEPNet) that achieves precise plant organ segmentation and stem-leaf phenotypic trait extraction. The adopted simple-but-effective network architecture and innovative modern operations, including a high-dimensional feature mapping strategy for preprocessing input points, a local feature extraction module based on inverted residual bottleneck block, and a cost-free attention block for spatial feature fusion, effectively implement multi-scale hierarchies and adaptively reduce computational overheads. Experimental results from cotton stem-leaf segmentation demonstrated that PEPNet not only presented approximately 2 × faster inference speed (9.59 ms) and throughput (146.32 plants per second) but also achieved competitive segmentation performance compared to other six state-of-the-art deep learning networks, namely PointNet++, DGCNN, CurveNet, Point Cloud Transformer, PointMLP, and SPoTr, achieving 95.99 %, 94.66 %, 95.32 %, and 91.31 % in Precision, Recall, F1-score, and mIoU, respectively. In transferability experiments with tomato and soybean plants, PEPNet achieved almost all the best metrics and significantly outperformed the second-best model (CurveNet). Furthermore, ablation study verified the optimal trade-off between efficiency and accuracy in this network. Any modifications to the modules could potentially disrupt the optimal trade-off. This work could contribute to reducing computational resources and annotation costs for applying segmentation methods in high-throughput phenotyping tasks. [ABSTRACT FROM AUTHOR]