首页期刊导航|Computers and Electronics in Agriculture
期刊信息/Journal information
Computers and Electronics in Agriculture
Elsevier Science Publishers
Computers and Electronics in Agriculture

Elsevier Science Publishers

0168-1699

Computers and Electronics in Agriculture/Journal Computers and Electronics in AgricultureSCIEIISTP
正式出版
收录年代

    An independent steering driving system to realize headland turning of unmanned tractors

    Yang Y.Zhang G.Chen Z.Wen X....
    11页
    查看更多>>摘要：? 2022Headland turning is one of the factors that affect the efficiency of automatic tractors. The tractor turns in headland following a curved trajectory based on the minimum turning radius and drives into the next row, in most cases. However, the efficiency of the tractor with this turning method is limited by the steering angle range of the front wheels. In this paper, an auxiliary wheel is designed to achieve efficient headland turning by equipping a 360° steering auxiliary wheel at the front of the tractor, which includes auxiliary wheel structure, hydraulic drive system and electro-hydraulic servo controller. The designed electro-hydraulic system can realize automatic steering, lifting and driving of the auxiliary wheel. Finally, field experiments are conducted to verify the superiority of the proposed auxiliary wheel turning method compared with the existing turning methods. Experimental results show that the auxiliary wheel turning method improved the efficiency of the automatic tractor in headland by 50%, 80% and 50% respectively in terms of time efficiency, travel distance and appropriation of space compared with the conventional turning method.
      原文链接:
    • NSTL
    • Elsevier

    Predicting individual apple tree yield using UAV multi-source remote sensing data and ensemble learning

    Chen R.Zhang C.Xu B.Zhu Y....
    15页
    查看更多>>摘要：? 2022 Elsevier B.V.As one of the world's most popular fruit, apple tree yield prediction before harvest plays an important role in optimizing orchard nutrition management, especially at the individual tree level. However, few studies focus on fruit-tree yield prediction with remote-sensing technology whereas most of them aim at field crops. Current fruits identifying and counting methods often fail to produce the expected result due to light and occlusion in complex orchard conditions. Since both the spectral and morphological characteristics of tree canopy can reflect the growth and development of fruit trees and are directly related to its potential yield. In this study, we develop a channel for automatic extraction of spectral and morphological features of apple trees using light detection and ranging (LiDAR) and multispectral imagery data from unmanned aerial vehicles. The contribution of spectral and morphological characteristics to the yield prediction of individual apple trees is discussed. With the combination of spectral and morphological features, an ensemble machine learning yield prediction model was developed by combining two widely used basic learners: support vector regression (SVR) and K-nearest neighbor (KNN). Then through extrapolating the ensemble model, the yield map was produced at the orchard level and individual tree level, respectively. The results show that the data processing channels developed in this study can accurately extract the morphological and spectral features of individual apple trees. Three features (Crown Volume 1, Ratio Vegetation Index, and CPA1) contribute most in apple tree yield prediction. The ensemble learning model outperforms all base learners with R2 = 0.813 for the validation and 0.758 for the test when using the selected three features. This study thus provides a practical example of predicting the yield of individual apple trees based on multi-source remote-sensing data and ensemble learning.
      原文链接:
    • NSTL
    • Elsevier

    CFD simulation and measurement of the downwash airflow of a quadrotor plant protection UAV during operation

    Zhu Y.Guo Q.Tang Y.Zhu X....
    11页
    查看更多>>摘要：? 2022 Elsevier B.V.To improve the effectiveness of unmanned aerial vehicle (UAV) spray, it is necessary to clarify the flow characteristics and penetration inside the canopy of the downwash airflow of UAV. This study used a three-dimensional computational fluid dynamic (CFD) model for the flow characteristics and velocity distribution of the downwash airflow generated by a quadrotor plant protection UAV at different flight velocities (1 ～ 5 m/s) was established. The measurement tests for airflow velocity distribution inside the canopies were conducted. The results indicated that the spiral vortices appeared below the rotors and then evolved into inclined horseshoe vortices. The horseshoe vortices at the flight velocity of 5 m/s exceeded the flight height of the UAV and tended to cause droplet drift. When the flight velocity was larger than 4 m/s, it was difficult for the horseshoe vortex to enter the bottom of the canopy, which was not conducive to improving the droplet penetration performance. The diffusion and attenuation of the horseshoe vortex during the flow process led to an increase and then a decrease in the velocity distribution area in the Z-direction of airflow. In order to improve the droplet penetration performance, the optimal operating distance range between UAV and canopy was given when the flight velocity was less than 3 m/s based on the variation law of the velocity distribution in the Z-direction. The validation results showed that the simulated and measured values maintained the same trend.
      原文链接:
    • NSTL
    • Elsevier

    Data synchronization for gas emission measurements from dairy cattle: A matched filter approach

    Milkevych V.Michelle Villumsen T.Lovendahl P.Sahana G....
    12页
    查看更多>>摘要：? 2022 The Author(s)Methane emissions from enteric fermentation in dairy cows make a substantial contribution to the greenhouse gas problem. Sniffer approaches represent a popular low-cost, mobile technology for cow gas emission measurement in industrial installations owing to their providing reliable emission estimates. Despite its advantages, the sniffer approach often yields data inconsistencies during automated acquisition due to unsynchronized clocks between sniffers and associated automatic milking machines, leading to uncertainty in the linking of each animal's data with methane emission records. Given the constantly growing demand for large-scale methane emission measurements for genetic studies, sniffer installations are expected to increase, making the need for an efficient solution to the data synchronization issue prescient. A novel approach for handling the synchronization problem was developed in this study based on matched filter theory. The approach was verified on gas emission data from multiple commercial dairy farms in Denmark. The results were analyzed and discussed in terms of accuracy and general characteristics of computational performance irrespective of a specific software implementation. The present findings support the conclusion that the presented matched filter-based approach is robust, applicable to the problem of cattle gas emission data synchronization, and convenient for automated data processing.
      原文链接:
    • NSTL
    • Elsevier

    Cotton canopy airflow simulation and velocity attenuation model based upon 3D phenotype and stratified sub-regional porous medium

    Cui H.Liu X.Yuan J.Liu Y....
    14页
    查看更多>>摘要：? 2022 Elsevier B.V.Airflow assisted with sprayers can change not only the canopy's porosity, but also allow the droplets to enter its interior. It is important to depict the airflow velocity distribution in a dense crop canopy quantitatively to optimize the goal of uniform deposition of droplets and full coverage within the canopy. However, there is yet no effective way to do so. The contribution of this paper is to propose a digital method to predict airflow distribution in the canopy using computational fluid dynamics (CFD) and machine-learning (ML) methods. Firstly, a virtual, simplified 3D cotton plant is generated based upon the phenotypic traits and the Logistic growth function to describe such canopy architecture parameters as leaf area density (LAD) and porosity quantitatively at any position in the canopy. Secondly, to solve the problem of uneven distribution of airflow in the canopy, in the CFD model, the target area surrounded by the main stems of 4 adjacent 3D cotton plants is selected as a stratified sub-regional porous medium (SSPM), and the accuracy of the CFD simulations is compared to the indoor measurements. Thereafter, a range of ML algorithms are trained with the velocity dataset obtained from the CFD results under different spray operating parameters (initial air velocity, canopy depth, LAD, and porosity). The comparison of CFD simulations and actual airflow measurements shows that the mean normalized mean absolute errors (NMAEs) of the lower, middle, and upper layers are 17.38 %, 21.35 %, and 9.75 % respectively. The Random Forest method has greater prediction accuracy, with a Coefficient of Determination (R2) of 0.9882 and Root Mean Squared Error (RMSE) of 0.5071, which indicates excellent agreement with the CFD simulation results. Therefore, such a data-driven model of airflow velocity distribution in a dense crop canopy can be used to optimize the operation parameters and eliminate complex CFD simulations or tedious physical experiments.
      原文链接:
    • NSTL
    • Elsevier

    HSI-TransUNet: A transformer based semantic segmentation model for crop mapping from UAV hyperspectral imagery

    Niu B.Feng Q.Chen B.Yang J....
    15页
    查看更多>>摘要：? 2022 Elsevier B.V.UAV hyperspectral imagery (HSI) has the unique merits of both a very high spatial and spectral resolution, which provides a high-quality data source for automatic crop mapping. Recently, deep learning has been widely used in crop classification, however, the design of an accurate crop mapping model for HSI data still remains a challenging task. Therefore, this paper aims to propose a novel semantic segmentation model (HSI-TransUNet) for crop mapping, which could make full use of the abundant spatial and spectral information of UAV HSI data simultaneously. Specifically, the proposed HSI-TransUNet belongs to an improved version of TransUNet, and we have made four important modifications for HSI data. Firstly, a spectral-feature attention module is designed for spectral features aggregation in the encoder. Afterwards, a series of Transformer layers with residual connections are designed to learn global contextual features. In the decoder part, sub-pixel convolutions are adopted to avoid the chess-board effect in the segmentation results. Finally, we design a hybrid loss function to further refine the predictions for boundaries. Experiment results indicate that the proposed HSI-TransUNet has achieved good performance in crops identification with an overall accuracy of 86.05%. Ablation studies have been conducted to verify the effectiveness of each refined module in the HSI-TransUNet. Comparison experiments also show that HSI-TransUNet has outperformed several previous semantic segmentation models. The dataset in this paper, UAV-HSI-Crop, is publicly available. http://doi.org/10.57760/sciencedb.01898.
      原文链接:
    • NSTL
    • Elsevier

    Integrating an attention-based deep learning framework and the SAFY-V model for winter wheat yield estimation using time series SAR and optical data

    Wang P.Zhang Y.Tian H.Tansey K....
    13页
    查看更多>>摘要：? 2022Information on the spatial distribution of yields can be obtained over a large area by using remote sensing (RS) data. Combining Synthetic Aperture Radar (SAR), being sensitive to above ground biomass and soil moisture in all weather conditions, and optical data can improve the usability of RS data and provide a basis for pixel-based crop yield estimation (YE). In this study, an Upscaled Convolutional Gated Recurrent Unit model incorporated an attention mechanism (UpSc-AConvGRU model) was proposed to improve the estimation accuracy of the winter wheat growth parameter, Leaf Area Index (LAI). Gap filling the time series of optical data was done with backscatter coefficients, local incidence angles and polarimetric decomposition information from Sentinel-1 SAR imagery. The time series LAI estimated by the UpSc-AConvGRU model and Vegetation Temperature Condition Index (VTCI) retrieved from Sentinel-3 optical imagery were then used as state variables of the SAFY-V model to estimate winter wheat yield. The results showed that the proposed UpSc-AConvGRU model incorporated the Convolutional Block Attention Module (CBAM) can effectively improve the accuracy of LAI estimation, with RMSEs ranging from 0.413 to 0.699 m2 m2 for LAI estimated within main growth stages (MGSs) of winter wheat. The correlation between estimated LAI and Sentinel-3 retrieved LAI was generally higher at irrigated farmland compared to rain-fed farmland. The estimated LAI was closest to Sentinel-3 retrieved LAI at the green-up and late heading-filling stage of winter wheat, followed by the jointing and early heading-filling stage, and finally the milk maturity stage. There was good agreement between the SAFY-V model estimated and field measured winter wheat yields (R2 = 0.546, RMSE = 0.757 t ha?1), and the estimated yields at the pixel scale in the Guanzhong Plain, PR China were satisfactory. This study combined deep learning and crop growth modeling, proposed a new pixel scale winter wheat YE method.
      原文链接:
    • NSTL
    • Elsevier

    A novel labeling strategy to improve apple seedling segmentation using BlendMask for online grading

    Suo R.Fu L.He L.Li G....
    9页
    查看更多>>摘要：? 2022 Elsevier B.V.A large number of apple seedlings are planted in orchards each year, where accurate and fast seedling grading to ensure their quality before planting has become a crucial problem. However, seedling grading by manual measurement of morphological indicators is laborious and inaccurate, and it's thus highly desirable to be replaced by machine vision. Seedling segmentation is one of the key steps of measuring morphological indicators and grading by machine vision. Therefore, a segmentation method of apple seedlings based on BlendMask with ResNet-101 to do transfer learning was proposed. A total of 450 original images were captured with Azure Kinect DK sensor. Root, rootstock, graft union, and scion of apple seedlings were labeled using a novel labeling strategy, which probably affect segmentation of thin and long objects. Scion was labeled with three different strategies, namely whole labeling (WL), segmental labeling (SL), and segmental-end-merge labeling (SEML). Results showed that the most suitable strategy was the SL for scion, which obtained a mean average precision of 91.2 % and the highest mean intersection over union of 79.3 % in the three labeling strategies. The average precisions of root, rootstock, graft union, and scion with the SL were 98.9 %, 89.3 %, 90.6 %, and 85.6 %, respectively. Intersection over unions of root, rootstock, graft union, and scion by the SL were 87.2 %, 75.8 %, 69.3 %, and 84.9 %, respectively. And it cost about 285 ms on average to process an image with resolution 3840 × 2160 pixels. The above results illustrated that the SL strategy is conducive to improve segmentation precision of thin and long objects. Moreover, apple seedlings can be effectively segmented, which is beneficial for the machine vision to measure morphological indicators and grade.
      原文链接:
    • NSTL
    • Elsevier

    Improved greenhouse self-propelled precision spraying machine—Multiple height and level (MHL) control

    Fu Q.Li X.Zhang G.Ma Y....
    13页
    查看更多>>摘要：? 2022 Elsevier B.V.In the research, an improved greenhouse self-propelled precision sprayer was designed. The traditional greenhouse spraying machine mainly relies on manual or single height and level operation, but in most cases, this kind of operation method cannot fully meet the requirements of spray. To address this issue, in earlier work an optimization framework was developed, where in step one a fast and accurate identification method for crop disease severity, and in step two design and control a Multiple Height and Level (MHL) rack and image recognition, remote control technology and visible light spectrum information and other configurations to achieve the expected spray needs. Here the aims are: 1) to develop step two; 2) to illustrate the potential precision spraying value and cost savings of both steps by comparing the optimization results with real operating data from traditional greenhouse self-propelled sprayers, as a benchmark. The analysis results show that the improved greenhouse self-propelled precision sprayer has higher uniformity and better stability than the traditional greenhouse self-propelled sprayer.
      原文链接:
    • NSTL
    • Elsevier

    Segmentation algorithm for overlap recognition of seedling lettuce and weeds based on SVM and image blocking

    Zhang L.Zhang Z.Wu C.Sun L....
    10页
    查看更多>>摘要：? 2022For the problem of a low recognition rate and shape feature failure caused by overlapping seedlings and weeds during the development of an intelligent lettuce weeding robot, a method to identify seedling lettuce and weeds based on an image block and support vector machine (SVM) is proposed, which realizes their precise identification and boundary segmentation. The a* channel is used to grayscale the collected image. The Otsu and morphological methods are selected to extract all the green targets in the image. The connected component analysis method is applied to label the green targets with regions of interest (ROIs), and those with pixel areas larger than the area threshold are normalized to 256 × 256 pixels. The image blocking technique is introduced to separately aliquot the normalized ROI, with block sizes of 16 × 16, 32 × 32, and 64 × 64 pixels. On this basis, the image sub-blocks are manually labeled, block by block, to extract three texture features: histogram of oriented gradient (HOG), local binary pattern (LBP), and gray-level co-occurrence matrix (GLCM). With the accuracy of fivefold cross-validation as the optimization objective, a genetic algorithm (GA) is used to optimize the SVM penalty and kernel parameters of 21 groups of research objects (one block size has three texture features, which are arbitrarily combined to form seven research objects, with a total of three block sizes). We compare the recognition performance of the SVM, RF, KNN, and GA-SVM classifiers in a single feature and a combination of fusion strategies through comparative analysis. When the block size is 32 × 32 pixels, the fusion of LBP and GLCM features under the GA-SVM classifier has the highest accuracy, and the optimal SVM model for the identification of lettuce and weeds in the seedling stage is obtained. For the misidentified image sub-blocks in optimization model recognition, an image block reconstruction method based on the comparison of the center point and eight-neighbor label value is proposed, and this is combined with the proportion of image blocks of two labels for comprehensive judgment. The center point label value is reconstructed to the improve recognition accuracy. Experimental results show that the average precision, recall, and F1 score of the proposed method are 0.9473, 0.9529, and 0.9498, respectively, and those of images without overlapping leaves can all reach 1, thus providing a theoretical basis for crop recognition and segmentation.
      原文链接:
    • NSTL
    • Elsevier