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

    A general technique for the estimation of farm animal body part weights from CT scans and its applications in a rabbit breeding program

    Csoka A.Petnehazy O.Repa I.Donko T....
    10页
    查看更多>>摘要：? 2022 Elsevier B.V.Various applications of farm animal imaging are based on estimating the weight of specific body parts and cuts from CT images of animals. In many cases, the complexity of the problem is increased by the enormous variability in body postures in the CT images due to scanning unsedated live animals. In this work, we propose a general and robust approach to estimate the weights of sections and body parts from CT images of (possibly) live animals. We adapt the segmentation by elastic registration and joint feature and model selection for the regression component to cope with the large number of features and small number of samples. The proposed technique is evaluated and illustrated by real applications in rabbit breeding programs. The r2 values are 12% higher than those of previous techniques and methods used in the breeding selection programs. The proposed technique is easily transferable to similar problems and is therefore made available in an open source software package for the benefit of the community.
      原文链接:
    • NSTL
    • Elsevier

    The importance of modeling the effects of trend and anisotropy on soil fertility maps

    Oliveira A.L.G.Lima J.P.Brasco T.L.Amaral L.R....
    11页
    查看更多>>摘要：? 2022 Elsevier B.V.Precision Agriculture (PA) commonly uses interpolation to generate maps for site-specific management. Semivariogram modeling with kriging interpolation considerers several parameters such as trend and anisotropy, which require proper estimation to return reliable maps. Often present in agricultural fields, these directional effects can also account for machine traffic and crop/soil management. Despite modeling trend and anisotropy being desirable for creating soil fertility maps, these effects are often disregarded during semivariogram modeling. Hence, this study evaluates whether semivariogram modeling considering anisotropy and trend influences the improvement of maps used in precision agriculture. Predicted performance and trends identified in the data when modeling anisotropy were evaluated considering two sampling grid densities, using the method of moments (MoM) and restricted maximum likelihood (REML) to estimate semivariogram parameters. Different levels of trend and anisotropy were tested on four types of virtual fields with 100 repetitions, and two experimental fields. Data were processed in an automated manner for virtual field generation, sampling extraction, semivariogram modeling, kriging, and cross-validation. Metrics were then subjected to bootstrapping and the differences were compared using confidence intervals. Results indicate that modeling directional effects improved the accuracy of kriging-generated maps. REML resulted in the best variability estimation in strong anisotropy, whereas MoM was more efficient in fields with weaker anisotropy. Modeling anisotropy was particularly useful in experimental fields, where trend was considered a function of spatial covariates. Consequently, semivariogram modeling must consider both directional effects to provide accurate soil fertility maps for precision agriculture.
      原文链接:
    • NSTL
    • Elsevier

    Power consumption analysis of the multi-edge toothed device for shredding the residual film and impurity mixture

    Liang R.Zhang B.Zhou P.Li Y....
    10页
    查看更多>>摘要：? 2022 Elsevier B.V.The theoretical model for the power consumption of the device that shreds a mixture of residual mulch film and impurities was constructed using the theoretical method, and the key factors affecting shredder power consumption were obtained. Through a central composite design, regression analysis and analysis of variance, a relational model was constructed among the number of multi-edge toothed cutters, the clearance between multi-edge toothed cutters and between multi-edge toothed cutters and stationary blades, the speed of high-speed toothed rollers, the speed difference between high-speed and low-speed toothed rollers, and the shredding power consumption. Under the same key parameter values of the cutting device, the error between the predicted value of the shredding power consumption obtained from the model and the measured value of the verification test was 9.76 %, which proved that the established cutting power consumption model had good reliability. The research results can provide a theoretical basis for the research on the technology and equipment development for shredding the residual film and impurity mixture.
      原文链接:
    • NSTL
    • Elsevier

    Data considerations for developing deep learning models for dairy applications: A simulation study on mastitis detection

    Naqvi S.A.Barkema H.W.Deardon R.King M.T.M....
    10页
    查看更多>>摘要：? 2022 Elsevier B.V.With growing adoption of precision dairy technologies, the use of big data is becoming increasingly common in the dairy industry. The speed at which data are generated has led to increased interest in developing detection and predictive models for animal health and disease events using real time records. When combining data from multiple sources, statistical methods exist to account for the underlying heterogeneity in data collected from commercial farms, although its impact on predictive models is not known. We investigated how 4 different issues commonly seen in these large datasets impact the performance of deep recurrent neural networks (RNNs) trained to detect the onset of clinical mastitis (CM) in dairy cows. Data were simulated by first sampling from real-world data and adding noise, then defining the association between predictor variables and CM while incorporating parameters to reflect underlying heterogeneity: 1) random effects to reflect unmeasured variability at the farm level (3 levels – none, moderate, high); 2) random effects to reflect unmeasured variability at the cow level (3 levels – none, moderate, high); 3) missed recording of CM cases (3 false-negative rates – 0.10, 0.25, 0.50); and 4) incomplete observations due to certain farms not having a somatic cell count sensor (SCC data missing vs SCC data included). At baseline (moderate farm and cow random effects; moderate misclassification; 42% herds with SCC sensor) the model achieved a sensitivity and specificity of 86% and 90% respectively. Higher levels of unmeasured variability at the farm and cow levels resulted in reduced model performance (sensitivity and specificity of 76% and 85% at the highest levels), indicating that data collection and feature selection should be informed by previous knowledge of the associations between the outcome and predictors when possible, and that model performance may be limited when predictors are selected only from routinely collected data. However, even when 50% of CM cases were incorrectly recorded as CM-negative, model performance did not decrease, demonstrating that deep RNNs are robust to the level of misclassification that would be typically encountered in dairy datasets. RNNs were also able to accurately detect CM onset even when a highly predictive variable, somatic cell count, was excluded from training and test data, but the models took longer to train. The effect of unmeasured variability on model performance demonstrates how predictors should be selected for RNNs, whereas RNNs appear to be very robust to misclassification in training data as well as missing variables. Researchers developing studies using deep learning should therefore focus their attention more on predictor selection than on reducing or standardizing outcome recording, since RNNs appear to be robust to the latter, while being more strongly impacted by the former.
      原文链接:
    • NSTL
    • Elsevier

    UAV-based chlorophyll content estimation by evaluating vegetation index responses under different crop coverages

    Qiao L.Tang W.Gao D.Zhao R....
    12页
    查看更多>>摘要：? 2022 Elsevier B.V.Efficiently estimating chlorophyll content is important in monitoring the photosynthesis capacity and growth status of maize canopy in precision agriculture management. Vegetation index (VI) easily obtained by proximal remote sensing has been used as a non-destructive and high-throughput way in crop monitoring, especially in chlorophyll estimation. However, the estimated results of the field chlorophyll content by VIs always face challenges from soil background inhibition and estimation stability under the dynamic changes of vegetation biomass. Thus, an unmanned aerial vehicle (UAV)-based chlorophyll content estimation was conducted by evaluating VI responses under different crop coverages. An analysis was conducted on 36 VIs under different crop coverage conditions to explore their response differences and robustness for chlorophyll estimation. This work focused on the three kinds of VIs named simple vegetation index, modified vegetation index, and functional vegetation index. In 2020, at the experimental station of Dryland Farming Institute of Hebei Academy of Agriculture and Forestry Sciences, UAV carrying multispectral sensor was used to collect visible and near-infrared images of the canopy at the jointing stage of maize under six fertilization levels to obtain VIs. After the UAV fled, ground calibration and sample collection were performed simultaneously, and chlorophyll content was measured. For data processing, correlation coefficient method (CCM) and maximal information coefficient (MIC) were first used to analyze the correlation response characteristics of VIs and chlorophyll content under three different coverage levels. The results showed that when the level of canopy coverage was increased, the linear correlation between VIs and chlorophyll content was substantially reduced. The MIC response indicating linear and non-linear combination relationship was more robust. In addition, the VIs obtained by UAV had a significant linear correlation with maize canopy chlorophyll under low (0.05–0.35) and medium (0.35–0.48) coverage, but an obvious non-linear correlation under high (0.48–0.75) coverage. Chlorophyll-sensitive parameters were then screened based on methods of CCM, MIC, and random frog method (RFM), respectively. Partial least squares regression (PLS) and random forest (RF) algorithms were used to establish the maize canopy chlorophyll content detection models. The findings showed that when Green minus red vegetation index (GMR), Red light normalized value (NRI), Normalized difference red edge (NDRE), Modified simple ratio with red edge (MSRREG), Enhanced Vegetation Index (EVI), Normalized red green difference vegetation index (NDIg), Normalized red blue difference vegetation index (NDIb), Soil-adjusted vegetation index (SAVI), Optimized soil-adjusted vegetation index with red edge (OSAVIREG), Soil-atmospherically resistant vegetation index (SARVI) were selected based on RFM as the optimal spectral variables, the chlorophyll content detection model constructed based on PLS had the least numbers of characteristic variables and the best model accuracy. The training set R2 and RMSE were 0.753 and 2.089 mg/L, respectively, and the verification set R2 and RMSE were 0.682 and 2.361 mg/L, respectively. Field chlorophyll content and detection error distribution maps were also drawn and combined with the distribution of fertilization management to provide support for the UAV monitoring of crop growth in the field and variable fertilization management decisions.
      原文链接:
    • NSTL
    • Elsevier

    Using EfficientNet and transfer learning for image-based diagnosis of nutrient deficiencies

    Espejo-Garcia B.Malounas I.Mylonas N.Kasimati A....
    11页
    查看更多>>摘要：? 2022 Elsevier B.V.Early diagnosis of nutrient deficiencies can play a major role in avoiding significant agricultural losses and increasing the final yield while preserving the environment through efficient fertilizer usage. In this work, we study how well nutrient deficiency symptoms can be recognized in RGB images by using deep neural networks and transfer learning. Two different datasets, presenting real-world conditions, were used for this purpose. The first one was the Deep Nutrient Deficiency for Sugar Beet (DND-SB) dataset, which contains 5648 images of sugar beets presenting nitrogen (N), phosphorous (P), and potassium (K) deficiencies, the omission of liming (Ca) and full fertilization. The second one, collected on the field for this research and currently publicly available, was a dataset combining different orange tree images with iron (Fe), potasssium (K), magnesium (Mg), and manganese (Mn) deficiencies. Image classification via fine-tuning with EfficientNetB4, whose original weights came from a noisy student training on ImageNet, obtained the best performances on both datasets with 98.65% and 98.52% Top-1 accuracies. Additionally, the Grad-CAM++ analysis showed that the models were performing an accurate analysis of the most relevant part inside the images. Finally, the use of agricultural transfer learning did not report improvement in the performances.
      原文链接:
    • NSTL
    • Elsevier

    Determining application volume of unmanned aerial spraying systems for cotton defoliation using remote sensing images

    Chen P.Xu W.Zhan Y.Yang W....
    11页
    查看更多>>摘要：? 2022 Elsevier B.V.The Unmanned aerial spraying systems (UASS), with a precise positioning system and convenient control system, have attracted more attention from researchers and the market. However, the UASS operation parameter setting still depends on the operator, and there is still a gap between the intelligent spraying. The upgrade of UASS from automation to intelligence requires a “scientific brain” to make spraying decisions. In this study, the UASS equipped with centrifugal nozzles was used to simulate defoliant spraying, combined with RGB and multi-spectral cameras to collect remote sensing images of the target area. The droplet distribution data were obtained through two years of field trials in two places. A droplet distribution prediction model based on the spray volume of UASS and the remote sensing spectral index that characterizes the cotton canopy structure is established by the BP neural network and Bayesian regularization training algorithm. The cotton defoliant was verified using this model and combined with NY/T 3213–2018 standard. The research results show that it is feasible to use remote sensing images to determine the application volume of UASSs for cotton defoliant. Compared with the conventional application rate and overspray, the defoliant spray based on the decision model can achieve the expected cotton defoliation effect and reduce the application volume.
      原文链接:
    • NSTL
    • Elsevier

    Modeling ammonia volatilization following urea and controlled-release urea application to paddy fields

    Shi X.Feng P.Hu K.Li X....
    10页
    查看更多>>摘要：? 2022 Elsevier B.V.Ammonia volatilization, a main pathway of nitrogen (N) loss in paddy fields, can easily lead to environmental problems, such as haze and water eutrophication. Field measurement of ammonia volatilization is time-consuming and expensive. Soil–crop system models can simulate the effect of different fertilizer management practices on NH3 volatilization and compensate for the limitations of field measurement. Therefore, the WHCNS (soil Water Heat Carbon Nitrogen Simulator) model was used to evaluate the feasibility of simulating NH3 volatilization in paddy fields under different N fertilizer types and application methods. The field experiment, including four treatments of farmer's practice (FP), basal application of urea in one dose (SBN), split application of urea in three doses (SPN), and basal application of controlled-release urea in one dose (CRU), was conducted from 2013 to 2014. All treatments were conducted with a fertilizer rate of 165 kg N ha?1, except for FP with a rate of 210 kg N ha?1. The measured dry matter weight, grain yield, N uptake, and NH3 flux were used to test the WHCNS model. The model evaluation indices of Nash–Sutcliffe efficiency (NSE) and index of agreement (d) were close to 1 for the measured variables, except for NSE of yield (0.16). The normalized root mean square error (NRMSE) of NH3 flux was slightly over 30%, but lower than 30% for the other items. Results showed that the model can successfully simulate rice yield and NH3 flux of urea and controlled-release urea under different application methods. The simulated annual average cumulative NH3 losses of FP, SBN, SPN, and CRU accounted for 18.9%, 17.8%, 20.6%, and 10.1% of their N inputs, respectively. Compared with the FP treatment, the ratios of NH3 losses of the SBN and CRU treatments were decreased by 1.2%-8.8%. Compared with the SBN treatment (the same rate of 165 kg N ha?1), SPN reduced the peak of NH3 flux but increased the cumulated NH3 losses and CRU significantly reduced the NH3 flux and cumulative NH3 losses. The application rate of controlled-release urea could be further optimized by refined model calculations to reduce NH3 volatilization and total N loss while maintaining yield.
      原文链接:
    • NSTL
    • Elsevier

    The spectralrao-monitoring Python package: A RAO's Q diversity index-based application for land-cover/land-use change detection in multifunctional agricultural areas

    Tassi A.Massetti A.Gil A.
    11页
    查看更多>>摘要：? 2022 Elsevier B.V.Monitoring multifunctional agricultural areas is paramount to ensure their cost-effective management. The remote sensing-based detection of land-cover/land-use (LCLU) changes and analysis of vegetation dynamics constitute a relevant indicator to support robust monitoring schemes, allowing the control of agri-environmental conditions and enforcing related measures and policies. The Rao's Q diversity index (RaoQ) is frequently used to measure functional diversity in ecology, thanks to the textural analysis of the environment. This paper aims to develop and provide an open-source Python application whose workflow may constitute a RaoQ-based LCLU change monitoring tool for multifunctional agricultural areas. Here, a use case is presented for detecting and mapping LCLU changes leveraging the free and open access Landsat 8 (L8) satellite data. The workflow is organized in four main stages: (1) data processing; (2) Normalized Difference Vegetation Index (NDVI) calculation; (3) RaoQ calculation; and (4) detection and mapping of LCLU changes through thresholding of RaoQ. Three methodological approaches were developed (RaoC – “classic” RaoQ; RaoMD – “multidimensional” RaoQ, and “classic + multidimensional” RaoQ) with overall accuracies ranging from 0.88 to 0.92. An example of an agri-environmental monitoring decision-support framework based on spectralrao-monitoring is presented. The application is easily reproducible, and the code is fully available and utilizable with other sensors at different resolutions to support monitoring other types of agricultural areas.
      原文链接:
    • NSTL
    • Elsevier

    Applicability evaluation of a demand-controlled ventilation system in livestock

    Shin H.Kwak Y.Huh J.-H.Jo S.-K....
    14页
    查看更多>>摘要：? 2022 Elsevier B.V.The distribution of agricultural and livestock products has been limited owing to the recent rapid population growth and the COVID-19 pandemic; this has led to an increase in the demand for food security. The livestock industry is interested in increasing the growth performance of livestock that has resulted in the need for a mechanical ventilation system that can create a comfortable indoor environment. In this study, the applicability of demand-controlled ventilation (DCV) to energy-efficient mechanical ventilation control in a pigsty was analyzed. To this end, an indoor temperature and CO2 concentration prediction model was developed, and the indoor environment and energy consumption behavior based on the application of DCV control were analyzed. As a result, when DCV control was applied, the energy consumption was smaller than that of the existing control method; however, when it was controlled in an hourly time step, the increase in indoor temperature was large, and several sections exceeded the maximum temperature. In addition, when it was controlled in 15-min time steps, the increase in indoor temperature and energy consumption decreased; however, it was not energy efficient on days with high-outdoor temperature and pig heat.
      原文链接:
    • NSTL
    • Elsevier