Automatic stomata recognition and measurement based on improved YOLO deep learning model and entropy rate superpixel algorithm

Measuring stomatal and guard cell metrics for plant physiology and growth using StoManager1

Automated guard cell detection and measurement are vital for understanding plant physiological performance and ecological functioning in global water and carbon cycles. Most current methods for measuring guard cells and stomata are laborious, time-consuming, prone to bias, and limited in scale. We developed StoManager1, a high-throughput tool utilizing geometrical, mathematical algorithms, and convolutional neural networks to automatically detect, count, and measure over 30 guard cell and stomatal metrics, including guard cell and stomatal area, length, width, stomatal aperture area/guard cell area, orientation, stomatal evenness, divergence, and aggregation index. Combined with leaf functional traits, some of these StoManager1-measured guard cell and stomatal metrics explained 90% and 82% of tree biomass and intrinsic water use efficiency (iWUE) variances in hardwoods, making them substantial factors in leaf physiology and tree growth. StoManager1 demonstrated exceptional precision and recall (mAP@0.5 over 0.96), effectively capturing diverse stomatal properties across over 100 species. StoManager1 facilitates the automation of measuring leaf stomatal and guard cells, enabling broader exploration of stomatal control in plant growth and adaptation to environmental stress and climate change. This has implications for global gross primary productivity (GPP) modeling and estimation, as integrating stomatal metrics can enhance predictions of plant growth and resource usage worldwide. Easily accessible open-source code and standalone Windows executable applications are available on a GitHub repository (https://github.com/JiaxinWang123/StoManager1) and Zenodo (https://doi.org/10.5281/zenodo.7686022).

Performance and morphology of several soybean varieties and responses to pests and diseases in South Sulawesi

Soybeans are a commodity that is widely grown by farmers in rainfed rice fields in South Sulawesi. One of the determining factors in increasing soybean productivity in South Sulawesi is the type of variety. The aim of this research was to determine the characteristics, morphology and response to pests and diseases in several soybean varieties planted in rainfed rice fields in South Sulawesi. This research was carried out in Allepolea Village, Maros Regency in 2022 using a Randomized Block Design with 13 treatments and 3 replications. Varieties tested as treatments include: 1) Derap-1, 2) Devon-2, 3) Deja-1, 4) Anjasmoro, 5) Dena-2, 6) Dena-1, 7) Gepak Kuning, 8) Grobogan, 9) Devon-1, 10) Dega-1, 11) Deja-2, 12) Demas-1, and 13) Detap-1. The results showed that of the 13 varieties tested, the highest height was found in Devon-2 (33.67 cm) and Detap-1 (31.67 cm) in the vegetative phase and in the generative phase in Detap-1 (75.53 cm) and Gepak Yellow (74.67 cm). The largest number of branches is in Dena-1 (3.13 branches). The highest nitrogen content was found in Devon-1 (12.64 m2 per g). The largest leaf area was Detap-1 (4.15 cm2) and Gepak Kuning (4.15 cm2). The highest number of stomata was in Dena-1 (42.80 μm) and Deja-1 (44.00 μm). The highest stomata width was found in Gepak Kuning (2.76 μm). The lowest level of leaf damage due to attacks by Valanga sp (Acrididae) occurred in Grobogan (6.89 %) and Dega-1 (7.35 %). The lowest level of pod damage due to Nezara viridula attack was in Devon-2 (3.56 %) and Dena-2 (3.64 %). The lowest level of leaf damage due to Phaedonia inclusa attack occurred in Dega-1 (4.37 %), Dena-2 (4, 12 %), and Grobogan (4.69 %). Seed damage due to Cercospora sp attack was lowest on Dena-2 (0.81 %). The highest seed yield was in Dena-2 (3.78 t ha-1) and the lowest in Anjasmoro (1.93 t ha-1) and Deja-2 (2.02 t ha-1).

Labeled temperate hardwood tree stomatal image datasets from seven taxa of Populus and 17 hardwood species

Machine learning (ML) algorithms have shown potential in automatically detecting and measuring stomata. However, ML algorithms require substantial data to efficiently train and optimize models, but their potential is restricted by the limited availability and quality of stomatal images. To overcome this obstacle, we have compiled a collection of around 11,000 unique images of temperate broadleaf angiosperm tree leaf stomata from various projects conducted between 2015 and 2022. The dataset includes over 7,000 images of 17 commonly encountered hardwood species, such as oak, maple, ash, elm, and hickory, and over 3,000 images of 55 genotypes from seven Populus taxa. Inner_guard_cell_walls and whole_stomata (stomatal aperture and guard cells) were labeled and had a corresponding YOLO label file that can be converted into other annotation formats. With the use of our dataset, users can (1) employ state-of-the-art machine learning models to identify, count, and quantify leaf stomata; (2) explore the diverse range of stomatal characteristics across different types of hardwood trees; and (3) develop new indices for measuring stomata.

A Test Method for Finding Early Dynamic Fracture of Rock: Using DIC and YOLOv5

Intelligent monitoring and early warning of rock mass failure is vital. To realize the early intelligent identification of dynamic fractures in the failure process of complex fractured rocks, 3D printing of the fracture network model was used to produce rock-like specimens containing 20 random joints. An algorithm for the early intelligent identification of dynamic fractures was proposed based on the YOLOv5 deep learning network model and DIC cloud. The results demonstrate an important relationship between the overall strength of the specimen with complex fractures and dynamic fracture propagation, and the overall specimen strength can be judged semi-quantitatively by counting dynamic fracture propagation. Before the initiation of each primary fracture, a strain concentration area appears, which indicates new fracture initiation. The dynamic evolution of primary fractures can be divided into four types: primary fractures, stress concentration areas, new fractures, and cross fractures. The cross fractures have the greatest impact on the overall strength of the specimen. The overall identification accuracy of the four types of fractures identified by the algorithm reached 88%, which shows that the method is fast, accurate, and effective for fracture identification and location, and classification of complex fractured rock masses.

Convolutional Neural Network for Object Detection in Garlic Root Cutting Equipment

Traditional manual garlic root cutting is inefficient and can cause food safety problems. To develop food processing equipment, a novel and accurate object detection method for garlic using deep learning—a convolutional neural network—is proposed in this study. The you-only-look-once (YOLO) algorithm, which is based on lightweight and transfer learning, is the most advanced computer vision method for single large object detection. To detect the bulb, the YOLOv2 model was modified using an inverted residual module and residual structure. The modified model was trained based on images of bulbs with varied brightness, surface attachment, and shape, which enabled sufficient learning of the detector. The optimum minibatches and epochs were obtained by comparing the test results of different training parameters. Research shows that IRM-YOLOv2 is superior to the SqueezeNet, ShuffleNet, and YOLOv2 models of classical neural networks, as well as the YOLOv3 and YOLOv4 algorithm models. The confidence score, average accuracy, deviation, standard deviation, detection time, and storage space of IRM-YOLOv2 were 0.98228, 99.2%, 2.819 pixels, 4.153, 0.0356 s, and 24.2 MB, respectively. In addition, this study provides an important reference for the application of the YOLO algorithm in food research.