Search

When monitoring crop growth using top-down images of the plant canopies, leaves in agricultural fields appear very dense and significantly overlap each other. Moreover, the image can be affected by external conditions such as background environment and light intensity, impacting the effectiveness of image segmentation. To address the challenge of segmenting dense and overlapping plant leaves under natural lighting conditions, this study employed a Bilayer Convolutional Network (BCNet) method for accurate leaf segmentation across various lighting environments. The major contributions of this study are as follows: (1) Utilized Fully Convolutional Object Detection (FCOS) for plant leaf detection, incorporating ResNet-50 with the Convolutional Block Attention Module (CBAM) and Feature Pyramid Network (FPN) to enhance Region of Interest (RoI) feature extraction from canopy top-view images. (2) Extracted the sub-region of the RoI based on the position of the detection box, using this region as input for the BCNet, ensuring precise segmentation. (3) Utilized instance segmentation of canopy top-view images using BCNet, improving segmentation accuracy. (4) Applied the Varifocal Loss Function to improve the classification loss function in FCOS, leading to better performance metrics. The experimental results on cucumber canopy top-view images captured in glass greenhouse and plastic greenhouse environments show that our method is highly effective. For cucumber leaves at different growth stages and under various lighting conditions, the Precision, Recall and Average Precision (AP) metrics for object recognition are 97%, 94% and 96.57%, respectively. For instance segmentation, the Precision, Recall and Average Precision (AP) metrics are 87%, 83% and 84.71%, respectively. Our algorithm outperforms commonly used deep learning algorithms such as Faster R-CNN, Mask R-CNN, YOLOv4 and PANet, showcasing its superior capability in complex agricultural settings. The results of this study demonstrate the potential of our method for accurate recognition and segmentation of highly overlapping leaves in diverse agricultural environments, significantly contributing to the application of deep learning algorithms in smart agriculture.

Accurate detection and reconstruction of branches aid the accuracy of harvesting robots and extraction of plant phenotypic information. However, the complex orchard background and twisting growing branches of vine fruit trees make this challenging. To solve these problems, this study adopted a Mask Region-based convolutional neural network (Mask R-CNN) architecture incorporating deformable convolution to segment branches in complex backgrounds. Based on the growth posture, a branch reconstruction algorithm with bidirectional sector search was proposed to adaptively reconstruct the segmented branches obtained by an improved model. The average precision, average recall, and F1 scores of the improved Mask R-CNN model for passion fruit branch detection were found to be 64.30%, 76.51%, and 69.88%, respectively, and the average running time on the test dataset was 0.75 s per image, which is better than the compared model. We randomly selected 40 images from the test dataset to evaluate the branch reconstruction. The branch reconstruction accuracy, average error, average relative error of reconstructed diameter, and mean intersection-over-union (mIoU) were 88.83%, 1.98 px, 7.98, and 83.44%, respectively. The average reconstruction time for a single image was 0.38 s. This would promise the proposed method to detect and reconstruct plant branches under complex orchard backgrounds.

Abstract Image hashing is an efficient technology for processing digital images and has been successfully used in image copy detection, image retrieval, image authentication, image quality assessment, and so on. In this paper, we design a new image hashing with compressed sensing (CS) and ordinal measures. This hashing method uses a visual attention model called Itti model and Canny operator to construct an image representation, and exploits CS to extract compact features from the representation. Finally, the CS-based compact features are quantized via ordinal measures. L2 norm is used to judge similarity of hashes produced by the proposed hashing method. Experiments about robustness validation, discrimination test, block size discussion, selection of visual attention model, selection of quantization scheme, and effectiveness of the use of ordinal measures are conducted to verify performances of the proposed hashing method. Comparisons with some state-of-the-art algorithms are also carried out. The results illustrate that the proposed hashing method outperforms some compared algorithms in classification according to ROC (receiver operating characteristic) graph.

The extraction and analysis of plant phenotypic characteristics are critical issues for many precision agriculture applications. An improved YOLOv5 model was proposed in this study for accurate node detection and internode length estimation of crops by using an end-to-end approach. In this improved YOLOv5, a feature extraction module was added in front of each detection head, and the bounding box loss function used in the original network of YOLOv5 was replaced by the SIoU bounding box loss function. The results of the experiments on three different crops (chili, eggplant, and tomato) showed that the improved YOLOv5 reached 90.5% AP (average precision) and the average detection time was 0.019 s per image. The average error of the internode length estimation was 41.3 pixels, and the relative error was 7.36%. Compared with the original YOLOv5, the improved YOLOv5 had an average error reduction of 5.84 pixels and a relative error reduction of 1.61%.

Accurate structural phenotyping analysis is essential to understand plant architectural adaptation strategy to environment change. The aim of this study was to analyze leaf arrangement and geometry influenced by azimuthally generated light gradient; and to simulate static and heterogeneous cucumber canopies using regression equations by considering more geometric parameters. Three continuous measurements of structural organ parameters were obtained to fit the organ initiation and expansion curves. Four measurements with three density treatments were obtained to validate model accuracy. To describe leaf distribution and orientation characteristics in more detail, azimuth and elevation models were introduced into canopy structure modelling. Leaf distribution frequency was simulated based on leaf area index and solar elevation angle while leaf elevation was simulated based on leaf azimuth and acropetal phytomer number. This study provides an important basis for structural phenotyping analysis of cucumber canopy, which is essential for more accurate functional-structural modelling in the future. Keywords: Phenotyping, Structure, Heterogeneity, Model calibration, Azimuth, Elevation

Precise canopy management is critical in vineyards for premium wine production because maximum crop load does not guarantee the best economic return for wine producers. The growers keep track of the number of grape bunches during the entire growing season for optimizing crop load per vine. Manual counting of grape bunches can be highly labor-intensive and error prone. Thus, an integrated, novel detection model, Swin-transformer-YOLOv5, was proposed for real-time wine grape bunch detection. The research was conducted on two varieties of Chardonnay and Merlot from July to September 2019. The performance of Swin-T-YOLOv5 was compared against commonly used detectors. All models were comprehensively tested under different conditions, including two weather conditions, two berry maturity stages, and three sunlight intensities. The proposed Swin-T-YOLOv5 outperformed others for grape bunch detection, with mean average precision (mAP) of up to 97% and F1-score of 0.89 on cloudy days. This mAP was ~44%, 18%, 14%, and 4% greater than Faster R-CNN, YOLOv3, YOLOv4, and YOLOv5, respectively. Swin-T-YOLOv5 achieved an R2 of 0.91 and RMSE of 2.4 (number of grape bunches) compared with the ground truth on Chardonnay. Swin-T-YOLOv5 can serve as a reliable digital tool to help growers perform precision canopy management in vineyards.

The accuracy detection of individual citrus fruits in a citrus orchard environments is one of the key steps in realizing precision agriculture applications such as yield estimation, fruit thinning, and mechanical harvesting. This study proposes an improved object detection YOLOv5 model to achieve accurate the identification and counting of citrus fruits in an orchard environment. First, the latest visual attention mechanism coordinated attention module (CA) was inserted into an improved backbone network to focus on fruit-dense regions to recognize small target fruits. Second, an efficient two-way cross-scale connection and weighted feature fusion BiFPN in the neck network were used to replace the PANet multiscale feature fusion network, giving effective feature corresponding weights to fully fuse the high-level and bottom-level features. Finally, the varifocal loss function was used to calculate the model loss for better model training results. The results of the experiments on four varieties of citrus trees showed that our improved model proposed to this study could effectively identify dense small citrus fruits. Specifically, the recognized AP (average precision) reached 98.4%, and the average recognition time was 0.019 s per image. Compared with the original YOLOv5 (including deferent variants of n, s, m, l, and x), the increase in the average accuracy precision of the improved YOLOv5 ranged from 7.5% to 0.8% while maintaining similar average inference time. Four different citrus varieties were also tested to evaluate the generalization performance of the improved model. The method can be further used as a part in a vision system to provide technical support for the real-time and accurate detection of multiple fruit targets during mechanical picking in citrus orchards.

Plant photosynthesis and biomass production are associated with the amount of intercepted light, especially the light distribution inside the canopy. Three virtual canopies (n = 80, 3.25 plants/m2) were constructed based on average leaf size of the digitized plant structures: ‘small leaf’ (98.1 cm2), ‘medium leaf’ (163.0 cm2) and ‘big leaf’ (241.6 cm2). The ratios of diffuse light were set in three gradients (27.8%, 48.7%, 89.6%). The simulations of light interception were conducted under different ratios of diffuse light, before and after the normalization of incident radiation. With 226.1% more diffuse light, the result of light interception could increase by 34.4%. However, the 56.8% of reduced radiation caused by the increased proportion of diffuse light inhibited the advantage of diffuse light in terms of a 26.8% reduction in light interception. The big-leaf canopy had more mutual shading effects, but its larger leaf area intercepted 56.2% more light than the small-leaf canopy under the same light conditions. The small-leaf canopy showed higher efficiency in light penetration and higher light interception per unit of leaf area. The study implied the 3D structural model, an effective tool for quantitative analysis of the interaction between light and plant canopy structure.

The leaf is the organ that is crucial for photosynthesis and the production of nutrients in plants; as such, the number of leaves is one of the key indicators with which to describe the development and growth of a canopy. The irregular shape and distribution of the blades, as well as the effect of natural light, make the segmentation and detection process of the blades difficult. The inaccurate acquisition of plant phenotypic parameters may affect the subsequent judgment of crop growth status and crop yield. To address the challenge in counting dense and overlapped plant leaves under natural environments, we proposed an improved deep-learning-based object detection algorithm by merging a space-to-depth module, a Convolutional Block Attention Module (CBAM) and Atrous Spatial Pyramid Pooling (ASPP) into the network, and applying the smoothL1 function to improve the loss function of object prediction. We evaluated our method on images of five different plant species collected under indoor and outdoor environments. The experimental results demonstrated that our algorithm which counts dense leaves improved average detection accuracy of 85% to 96%. Our algorithm also showed better performance in both detection accuracy and time consumption compared to other state-of-the-art object detection algorithms.

A venation skeleton-driven method for modeling and animating plant leaf wilting is presented. The proposed method includes five principal processes. Firstly, a three-dimensional leaf skeleton is constructed from a leaf image, and the leaf skeleton is further used to generate a detailed mesh for the leaf surface. Then a venation skeleton is generated interactively from the leaf skeleton. Each vein in the venation skeleton consists of a segmented vertices string. Thirdly, each vertex in the leaf mesh is banded to the nearest vertex in the venation skeleton. We then deform the venation skeleton by controlling the movement of each vertex in the venation skeleton by rotating it around a fixed vector. Finally, the leaf mesh is mapped to the deformed venation skeleton, as such the deformation of the mesh follows the deformation of the venation skeleton. The proposed techniques have been applied to simulate plant leaf surface deformation resulted from biological responses of plant wilting.

Abstract In this study, we introduce a highly effective non-metallic iodine single-atom catalyst (SAC), referred to as I-NC, which is strategically confined within a nitrogen-doped carbon (NC) scaffold. This configuration features a distinctive C-I coordination that optimizes the electronic structure of the nitrogen-adjacent carbon sites. As a result, this arrangement enhances electron transfer from peroxymonosulfate (PMS) to the active sites, particularly the electron-deficient carbon. This electron transfer is followed by a deprotonation process that generates the peroxymonosulfate radical (SO5 •−). Subsequently, the SO5 •− radical undergoes a disproportionation reaction, leading to the production of singlet oxygen (1O2). Furthermore, the energy barrier for the rate-limiting step of SO5 •− generation in I-NC is significantly lower at 1.45 eV, compared to 1.65 eV in the NC scaffold. This reduction in energy barrier effectively overcomes kinetic obstacles, thereby facilitating an enhanced generation of 1O2. Consequently, the I-NC catalyst exhibits remarkable catalytic efficiency and unmatched reactivity for PMS activation. This leads to a significantly accelerated degradation of pollutants, evidenced by a relatively high observed kinetic rate constant (k obs ~ 0.436 min− 1) compared to other metallic SACs. This study offers valuable insights into the rational design of effective non-metallic SACs, showcasing their promising potential for Fenton-like reactions in water treatment applications.

Antimony pollution poses a significant threat to both the ecological environment and the health of people living in mining regions. Using organic fertilizers is an efficient approach for the remediation of heavy metal contamination in soil. This study aimed to explore how food waste organic fertilizer (FF) can remediate antimony-contaminated soil and the associated rhizosphere microbial response mechanism. The analysis of soil physicochemical properties revealed that the application of FF notably reduced bulk density (from 1.57 to 1.08 g cm−3), enhanced salinization levels, and increased the content of organic matter, available nitrogen, phosphorus, and potassium (p

Abstract It is still a challenge to construct single‐atom level reduction and oxidation sites in single‐component photocatalyst by manipulating coordination configuration for photocatalytic water splitting. Herein, the atomically dispersed asymmetric configuration of six‐coordinated Co‐S2O4 (two exposed S atoms, two OH groups, and two Co─O─Zn bonds) suspending on ZnIn2S4 nanosheets verified by combining experimental analysis with theoretical calculation, is applied into photocatalytic water splitting. The Co‐S2O4 site immobilized by Vs acts as oxidation sites to guide electrons transferring to neighboring independent S atom, achieving efficient separation of reduction and oxidation sites. It is worth mentioning that stabilized Co‐S2O4 configuration show dynamic structure evolution to highly active Co‐S1O4 configuration (one exposed S atom, one OH group, and three Co─O─Zn bonds) in reaction, which lowers energy barrier of transition state for H2O activization. Ultimately, the optimized photocatalyst exhibits excellent photocatalytic activity for water splitting (H2: 80.13 µmol g−1 h−1, O2: 37.81 µmol g−1 h−1) and outstanding stability than that of multicomponent photocatalysts due to dynamic and reversible evolution between stable Co‐S2O4 configuration and active Co‐S1O4 configuration. This work demonstrates new cognitions on immobilized strategy through vacancy inducing, manipulating coordination configuration, and dynamic evolution mechanism of single‐atom level catalytic site in photocatalytic water splitting.

Abstract Extreme fast charging of Ampere-hour (Ah)-scale electrochemical energy storage devices targeting charging times of less than 10 minutes are desired to increase widespread adoption. However, this metric is difficult to achieve in conventional Li-ion batteries due to their inherent reaction mechanism and safety hazards at high current densities. In this work, we report 1 Ah soft-package potassium-ion hybrid supercapacitors (PIHCs), which combine the merits of high-energy density of battery-type negative electrodes and high-power density of capacitor-type positive electrodes. The PIHC consists of a defect-rich, high specific surface area N-doped carbon nanotube-based positive electrode, MnO quantum dots inlaid spacing-expanded carbon nanotube-based negative electrode, carbonate-based non-aqueous electrolyte, and a binder- and current collector-free cell design. Through the optimization of the cell configuration, electrodes, and electrolyte, the full cells (1 Ah) exhibit a cell voltage up to 4.8 V, high full-cell level specific energy of 140 Wh kg−1 (based on the whole mass of device) with a full charge of 6 minutes. An 88% capacity retention after 200 cycles at 10 C (10 A) and a voltage retention of 99% at 25 ± 1 °C are also demonstrated.