Search

The agriculture industry is one that is highly resource- and labour-intensive. As such, farmers are increasingly turning to technology and automation to address this issue. However, agricultural robots are far too complicated, slow, and costly to be made publicly available. As a result, the agriculture sector still lags behind in integrating modern technologies. This research paper details the development of a low-cost agricultural robot for spraying fertilizers and pesticides in agriculture fields as well as for general crop monitoring. The prototype system is a two-wheeled robot that consists of a mobile base, a spraying mechanism, a wireless controller for controlling the robot movement, and a camera for crop health and growth monitoring as well as detecting the presence of pests in the agriculture field. Tests conducted on the prototype system show that while the productivity of the robot in terms of crop coverage is slightly lower than a human worker, the labour cost savings afforded by the agricultural robot prototype is much greater as it functions completely in an autonomous mode and only requires the operator to control the robot when placing it at the start of the crop path. Furthermore, the prototype system also provides greater resource savings and reduction in the contamination of underground water sources due to leeching process, thus achieving precision agriculture goals. Lastly, the excellent battery life of the prototype system ensures that there will be no increase in the operation times and reduction in the efficiency of the fertilizer and pesticide spraying process due to the recharging times when replacing human workers. Future recommendations include making the agricultural robot fully autonomous, using either a rail- or line-following system, to further reduce the labour requirements and costs.

International audience; Portable mid-infrared spectroscopy (pMIRS) combined with machine learning was used to predict selected parameters for soil organic carbon (SOC) storage. In particular, SOC, soil inorganic C (SIC), hot-water extractable C (hwC), clay and sand content were predicted for ten vineyards with varying geopedological settings. As a pre-test, spectra were collected from sieved and pressed tablets with 30 and 90 kPa respectively and compared to powdery samples in order to optimise sample preparation. Further, spectra from 30 kPa tablets were used to calibrate prediction models for a sample set (n = 540) of 10 vineyards with pronounced geopedological variation using Support Vector Machines (SVM). The calibrated SVM models performed well with R2 = 0.81–0.98 and RPIQ = 5.20–13.0 for all investigated parameters. Third, two years after the calibration samples, follow-up samples were collected from four of the vineyards. While the models performed excellent for hwC (R2 = 0.93), prediction accuracy for SOC was lower. Segmentation of the total dataset into SIC-free and SIC-containing samples resulted in better predictions of SOC of the first sampling period. For the prediction of the follow-up sampling dates, model performance could not be maintained. We conclude that pMIRS-SVM calibrations are suited for the prediction of parameters related to soil C storage under varying geopedological conditions and may provide potential for future C monitoring. Extending the database with additional samples from geopedological scenarios not included in this dataset may strengthen model robustness and help to evaluate effects of SIC content on model performance.

In precision agriculture, the accurate segmentation of crops and weeds in agronomic images has always been the center of attention. Many methods have been proposed but still the clean and sharp segmentation of crops and weeds is a challenging issue for the images with a high presence of weeds. This work proposes a segmentation method based on the combination of semantic segmentation and K-means algorithms for the segmentation of crops and weeds in color images. Agronomic images of two different databases were used for the segmentation algorithms. Using the thresholding technique, everything except plants was removed from the images. Afterward, semantic segmentation was applied using U-net followed by the segmentation of crops and weeds using the K-means subtractive algorithm. The comparison of segmentation performance was made for the proposed method and K-Means clustering and superpixels algorithms. The proposed algorithm provided more accurate segmentation in comparison to other methods with the maximum accuracy of equivalent to 99.19%. Based on the confusion matrix, the true-positive and true-negative values were 0.995 2 and 0.898 5 representing the true classification rate of crops and weeds, respectively. The results indicated that the proposed method successfully provided accurate and convincing results for the segmentation of crops and weeds in the images with a complex presence of weeds.

In spite of the insight commonality may have concerning agrarian practice, fact is that nowadays agricultural science diligence is accurate, precise, data-driven, and vigorous than ever. The emanation of the technologies based on Internet of Things (IoT) has reformed nearly each industry like smart city, smart health, smart grid, smart home, including “smart agriculture or precision agriculture”. Applying machine learning using the IoT data analytics in agricultural sector will rise new benefits to increase the quantity and quality of production from the crop fields to meet the increasing food demand. Such world-shattering advancements are rocking the current agrarian approaches and generating novel and best chances besides a number of limitations. This paper climaxes the power and capability of computing techniques including internet of things, wireless sensor networks, data analytics and machine learning in agriculture. The paper proposed the prediction model of Apple disease in the apple orchards of Kashmir valley using data analytics and Machine learning in IoT system. Furthermore, a local survey was conducted to know from the farmers about the trending technologies and their effect in precision agriculture. Finally, the paper discusses the challenges faced when incorporating these technologies in the traditional farming approaches.

Precision agriculture seeks to optimize production processes by monitoring and analyzing environmental variables. For example, establishing farming actions on the crop requires analyzing variables such as temperature, ambient humidity, soil moisture, solar irradiance, and Rainfall. Although these signals might contain valuable information, it is vital to mix up the monitored signals and analyze them as a whole to provide more accurate information than analyzing the signals separately. Unfortunately, monitoring all these variables results in high costs. Hence it is necessary to establish an appropriate method that allows the infer variables behavior without the direct measurement of all of them. This paper introduces a multi-sensor data fusion technique, based on a sparse representation, to find the most straightforward and complete linear equation to predict and understand a particular variable behavior based on other monitored environmental variables measurements. Moreover, this approach aims to provide an interpretable model that allows understanding how these variables are combined to achieve such results. The fusion strategy explained in this manuscript follows a four-step process that includes 1. data cleaning, 2. redundant variable detection, 3. dictionary generation, and 4. sparse regression. The algorithm requires a target variable and two highly correlated signals. It is essential to point out that the developed method has no restrictions to specific variables. Consequently, it is possible to replicate this method for the semiautomatic prediction of multiple critical environmental variables. As a case study, this work used the SML2010 data set of the UCI machine learning repository to predicted the humidity's derivative trend function with an error rate lower than 17% and a mean absolute error lower than 6%. The experiment results show that even though sparse model predictions might not be the most accurate compared to those of linear regression (LR), support vector machine (SVM), and extreme learning machine (ELM) since it is not a black-box model, it guarantees greater interpretability of the problem.

Soil moisture has continued to occupy the forefront of several interdisciplinary debates globally, particularly in agriculture, climatology, hydrology and civil engineering for several decades. Climate smart and precision farming depend extensively on knowledge of the volumetric pattern and trend in water content of the soil particularly at the root zone. This study evaluated the temporal and spatial variation of soil moisture (0–40 cm depth) in Etsako West Local Government Area, Edo State, Nigeria. It used time series soil moisture data (1982–2019) sourced from the Famine Early Warning Systems Network (FEWS NET) and Land Data Assimilation System (FLDAS). Descriptive and inferential statistics as well as geographical information system (GIS) were used in data analysis. Results showed that Nitisols (Iyuku) had mean soil moisture of 9.0285 m3/m3, Lixisols (Ugieda) recorded 9.0574 m3/m3 whereas Acrisols (Idegun) was 9.1307 m3/m3. September emerged as month with peak soil moisture of 0.9 m3/m3 in Lixisols and Acrisols while January and February were the months with the lowest value of 0.5 m3/m3 across the three soil classes. Soil moisture in all the soil classes also exhibited a rising trend in the climatic period with annual increase of 0.002 m3/m3 in Nitisols, 0.002 m3/m3 in Lixisols and 0.001 m3/m3 in Acrisols. Temporal pattern of soil moisture was not statistically significance while spatial variation was statistical significant across the soil classes. In a spatial data-driven society, this study has therefore unlocks the potentials of GIS and remote sensing resources in assessing global environmental change indicators in data sparse regions.

The socio-economic factor analyses by using the Logit and Probit model have the limitation of representing random taste variation for the unobserved factors and the issues of temporally correlated errors respectively. In socio-economic factor analysis, the observed data are essential in the random distribution for the adequate representation of the random components associated with various factors and lead to poor prediction in the case of the Logit and Probit model. In this work, Extra- trees classifier machine learning model based socio-economic factors selection has been used and found capable to find out the socio-economic factors that contain relevant information to the target variable agricultural productivity. In addition to this, the voting classifiers ensemble learning approach is used for the prediction of agricultural productivity of respondents (farmers) from their socio-economic profiles. This proposed work has been evaluated by using the test case of analyzing the socio-economic factors of the farmers affecting agricultural productivity in Sambalpur District, in Odisha State, India. In this study, the farmers’ socio-economic data are collected by using structured interviews through questionnaires that are in line with standard Participatory Rural Appraisal. It is found that, the proposed approach of socio-economic factor identification is found efficient for finding the relationships between socio-economic factors and agricultural productivity and the proposed ensemble learning-based approach is efficient in terms of agricultural productivity prediction.

Rapid socio-economic changes in China, such as land conversion and urbanization, are creating new scopes for the application of precision agriculture (PA).An experiment to assess the economic benefits of two precision agriculture methods was applied for one year – precision seeding and precision seeding with land leveling. Whilst the results for this were positive, of itself it did not provide evidence of longer terms gains. The costs of land leveling are accrued in a single year but the benefits could carry over into subsequent years. Thus, in this case if the PA method provides carry over benefits to future years, the economic assessment would incorrectly assign all the costs to a single year of benefits i.e.the benefit-cost ratio would be underestimated. To gauge whether there was carry over benefits in future years we looked at NDVI and GUI as proxies for future year benefits. For the single year experiment, our results showed that: (1) Winter wheat yield was increased 23.2% through the integration of precision seeding and laser leveling technologies.(2) Both the single technology and the integrated technologies significant reduced the concentration of soil ammonium nitrogen at the depths of 60 cm; (3) The benefit/cost ratio's of the treatments exceeded that of the baseline by approximately 10% which translated to an increase of several hundred US$ per hectare. The NDVI analysis showed that the effect of laser land leveling could last to the next two years. When considering the multi-year impact of land leveling, the benefit/cost ratio of PSLL will increase to 23.5% and 22.9% with and without laser land leveling subsidies. Making clear the economic benefits of using PA technologies will likely promote application of the technologies in the region.

Prediction of pollination stages in oil palm plantations is an important research area in precision agriculture. Oil palm is known as the most efficient commercial crop with the capacity to fulfill the growing global demand for vegetable oil consumption. However, oil palm production dependence on pollination is experiencing issues with decreasing the actual yield. Consequently, alternative methods in commercial plantations such as human-assisted pollination and recently Wireless Sensor Network (WSN) have been deployed despite their high economic costs due to labor requirements. Oil palm assisted pollination requires precision, inspection, traceability, and validation processes in the field. Currently, all these processes are performed by humans that can be associated with false assumptions, uncertainty, and pollination latency. Therefore, Machine Learning (ML) approaches as a subset of the Artificial Intelligence (AI) domain provides efficient, cost-effective, and non-destructive solutions to determine these reproductive stages for future autonomous pollination system. Our goal was to reduce the variability of worker's performance in oil palms, using ML algorithms to make expert decisions and reduce the risk related to a transient workforce. This comparative empirical study examined and compared the performance of the Random Forest (RF) against k Nearest Neighbor (kNN) and Support Vector Machine (SVM) for classification of oil palm pre-anthesis and anthesis stages, dividing into four classes (1, 2, 3, and 4). These models were tested using thermal features (endogenous) individually and in combination with meteorological variables (exogenous). The performance of models is evaluated with specific measures of performance, such as overall user's and producer's accuracies and F-measure values derived from the confusion matrix. The results showed that the RF model produced better results with regard to average F-measure (88.6%, 71.83%), producer's accuracies (88.70%, 71.35%), and user's accuracies (88.27%, 72.36%) on test sets using exogenous + exogenous and endogenous feature sets, respectively. Among the three classifiers tested with two datasets, the RF method outperforms the other two popular algorithms, i.e., kNN and SVM with respect to accuracy and F-measure metrics. The results validated the significance of thermal parameters, provided valuable features to devise an intelligent pollination management system, and proved the feasibility of using the RF model for the classification of oil palms four stages of anthesis.

Semantic segmentation of remotely sensed imagery plays a critical role in many real-world applications, such as environmental change monitoring, precision agriculture, environmental protection, and economic assessment. Following rapid developments in sensor technologies, vast numbers of fine-resolution satellite and airborne remote sensing images are now available, for which semantic segmentation is potentially a valuable method. However, because of the rich complexity and heterogeneity of information provided with an ever-increasing spatial resolution, state-of-the-art deep learning algorithms commonly adopt complex network structures for segmentation, which often result in significant computational demand. Particularly, the frequently-used fully convolutional network (FCN) relies heavily on fine-grained spatial detail (fine spatial resolution) and contextual information (large receptive fields), both imposing high computational costs. This impedes the practical utility of FCN for real-world applications, especially those requiring real-time data processing. In this paper, we propose a novel Attentive Bilateral Contextual Network (ABCNet), a lightweight convolutional neural network (CNN) with a spatial path and a contextual path. Extensive experiments, including a comprehensive ablation study, demonstrate that ABCNet has strong discrimination capability with competitive accuracy compared with state-of-the-art benchmark methods while achieving significantly increased computational efficiency. Specifically, the proposed ABCNet achieves a 91.3% overall accuracy (OA) on the Potsdam test dataset and outperforms all lightweight benchmark methods significantly. The code is freely available at https://github.com/lironui/ABCNet .

In India, half the population depends on agriculture for a livelihood. Microbial diseases are a significant threat to food security, but their rapid identification remains difficult due to limited infrastructure. With AI, automatic detection of plant diseases from raw images is possible using deep learning and transfer learning. This paper aims to detect and classify Grapes and Mango leaf diseases, employing a dataset of 8,438 images of diseased and healthy leaves collected from the Plant Village dataset and acquired locally. The deep convolutional neural network (CNN) is trained to identify diseases or their absence. A pre-trained CNN architecture called AlexNet is modeled for automatic feature extraction and classification. The system is developed with MATLAB achieves an accuracy rate of the detection of 99% and 89% for Grape leaves and Mango leaves respectively. An app named "JIT CROPFIX" is developed to implement the same on an Android Smartphone.

Excess calcium (Ca) in soils of semi-arid and arid regions has negative effects on soil structure and chemical properties, which limits the crop root growth as well as the availability of soil water and nutrients. Quantifying the spatial variability of soil Ca contents may reveal factors influencing soil erosion and provide a basis for site-specific soil and crop management in semi-arid regions. This study sought to assess the spatial variability of soil Ca in relation to topography, hydraulic attributes, and soil types for precision soil and crop management in a 194-ha production field in the Southern High Plains of Texas, USA. Soils at four depth increments (0–2, 0–15, 15–30, and 30–60 cm) were sampled at 232 points in the spring of 2017. The Ca content of each sample was determined with a DP-6000 Delta Premium portable X-ray fluorescence (PXRF) spectrometer. Elevation data was obtained using a real-time kinematic GPS receiver with centimeter-level accuracy. A digital elevation model (DEM) was derived from the elevation data, and topographic and hydraulic attributes were generated from this DEM. A generalized least-squares model was then developed to assess the relationship between soil Ca contents of the four layers and the topographic and hydraulic attributes. Results showed that topographic attributes, especially slope and elevation, had a significant effect on soil Ca content at different depths (P

Holistic information systems for climate-smart agriculture demands the seamless integration of various categories of climate, meteorological and weather data. Any actor in the agricultural value chain may harness weather forecasts at the short and medium-range, local weather history, and prevailing climatic conditions, to inform decision-making. Weather is fundamental to many day-to-day operations, especially at farm-level, influencing decision-making at various spatial and temporal scales. Many operational decisions ideally require hyper-localized service provision. In practice, integrating weather information into decision-support services demands a comprehensive understanding of various categories of weather-related data, their genesis, as well as the specific standards and data formats used by the meteorological community. This paper considers the weather as a crucial context for the delivery of farm-level operational services in smart agriculture, highlighting critical issues for reflection by system designers during the service design and implementation phases.

On farm experimentation (OFE) has been a long-standing method for farmers to assess alternative management at scales relevant to their farming practices. Through the use of spatially distributed designs, whether simple strips or other ‘whole-of-block’ trials, OFE can provide information such as which treatment should be recommended at specific locations, and make important contributions to precision agriculture. However, when treatment response data sets become large, such as with tens of thousands of field observations that are readily collected using on-the-go sensors, existing geostatistical systems for analysing such experiments become computationally intensive, if not impossible. To enable farmers, or their consultants, to generate high-resolution treatment response and recommendation maps on their own computers within a reasonable time, we present a fast and adaptive local cokriging tool for non-colocated and non-stationary OFE data. It uses a spatially-varying neighbourhood radius. It has a graphical user interface accessible via QGIS, a free and open source software. The adaptive local cokriging is demonstrated on three OFE examples. It performs indistinguishably from global cokriging on a small data set, but for large data sets, for which global cokriging is impractical, it predicts significantly more accurately than spatial splines or sampling-based cokriging. It outperforms cokriging base on a fixed number of nearest neighbours when this fixed number is not carefully chosen.

In this paper, we propose a novel deep learning method based on a Convolutional Neural Network (CNN) that simultaneously detects and geolocates plantation-rows while counting its plants considering highly-dense plantation configurations. The experimental setup was evaluated in a cornfield with different growth stages and in a Citrus orchard. Both datasets characterize different plant density scenarios, locations, types of crops, sensors, and dates. A two-branch architecture was implemented in our CNN method, where the information obtained within the plantation-row is updated into the plant detection branch and retro-feed to the row branch; which are then refined by a Multi-Stage Refinement method. In the corn plantation datasets (with both growth phases, young and mature), our approach returned a mean absolute error (MAE) of 6.224 plants per image patch, a mean relative error (MRE) of 0.1038, precision and recall values of 0.856, and 0.905, respectively, and an F-measure equal to 0.876. These results were superior to the results from other deep networks (HRNet, Faster R-CNN, and RetinaNet) evaluated with the same task and dataset. For the plantation-row detection, our approach returned precision, recall, and F-measure scores of 0.913, 0.941, and 0.925, respectively. To test the robustness of our model with a different type of agriculture, we performed the same task in the citrus orchard dataset. It returned an MAE equal to 1.409 citrus-trees per patch, MRE of 0.0615, precision of 0.922, recall of 0.911, and F-measure of 0.965. For citrus plantation-row detection, our approach resulted in precision, recall, and F-measure scores equal to 0.965, 0.970, and 0.964, respectively. The proposed method achieved state-of-the-art performance for counting and geolocating plants and plant-rows in UAV images from different types of crops., Comment: 27 pages, 12 figures, 9 tables

Fast and accurate plant disease detection is critical to increasing agricultural productivity in a sustainable way. Traditionally, human experts have been relied upon to diagnose anomalies in plants caused by diseases, pests, nutritional deficiencies or extreme weather. However, this is expensive, time consuming and in some cases impractical. To counter these challenges, research into the use of image processing techniques for plant disease recognition has become a hot research topic. In this paper, we provide a comprehensive review of recent studies carried out in the area of crop pest and disease recognition using image processing and machine learning techniques. We hope that this work will be a valuable resource for researchers in this area of crop pest and disease recognition using image processing techniques. In particular, we concentrate on the use of RGB images owing to the low cost and high availability of digital RGB cameras. We report that recent efforts have focused on the use of deep learning instead of training shallow classifiers using hand-crafted features. Researchers have reported high recognition accuracies on particular datasets but in many cases, the performance of those systems deteriorated significantly when tested on different datasets or in field conditions. Nevertheless, progress made so far has been encouraging. Experimental results showing the leaf disease recognition performance of ten CNN architectures in terms of recognition accuracy, recall, precision, specificity, F1-score, training duration and storage requirements are also presented. Subsequently, recommendations are made on the most suitable architectures to deploy in conventional as well as mobile/embedded computing environments. We also discuss some of the unresolved challenges that need to be addressed in order to develop practical automatic plant disease recognition systems for use in field conditions.

Information about soil physico-chemical parameters plays an important role in precision farming. To examine the relationship among soil properties, pedometric mapping is essential and has been widely applied in agricultural activities. This paper is aimed at soil nutrient assessment by generating the geospatial distribution maps for Kestur, a village in Tumakuru district of Karnataka State, India. Using random sampling technique, a total of one hundred and sixty geo-coordinated surface (0–20 cm) soil samples were collected from different land cover regions including irrigated and dryland areas in Kestur village. The samples collected were analysed for pH, electrical conductivity (EC), organic carbon (OC), available nitrogen (N), phosphorus (P as P2O5), and potassium (K as K2O). Also, a structured questionnaire was used to collect data about demographic profile, farming practices, crops grown and fertilizers applied from each farmer under the study. The spatial distribution maps were generated for each soil property and were mapped using spatial variability by adopting Ordinary Kriging geostatistical method. Semivariogram models such as exponential for pH, spherical model for EC and OC and Gaussian for N, P2O5, and K2O were used based on the best-fit selection using least Mean Error (ME) and Root Mean Square Error (RMSE) values. The outcome of the study proved that the geostatistical methods can be utilized to develop the spatial distribution maps thereby resulting in a cost effective solution to soil fertility management.

The tendencies of digital transformation of agriculture towards precision production of agricultural products using Earth remote sensing data, which have become one of the main drivers of the rapid development of the industry around the world, are considered. It is shown that these technologies are currently evolving from the digitalization of individual operations to the digitalization of an interconnected complex with the integration of all operations, including operations of related industries. An analysis of the problems of achieving effective implementation of digital technologies in Russia is given, as well as ways of solving them on the basis of complex, integration technologies. The transformation of agriculture into industrial production due to its digital transformation requires the development of methods for evaluating the economic efficiency of investing a new asset - precision farming technology and remote sensing of the Earth. The article presents the author’s mathematical model for evaluating the economic efficiency of using these technologies. Possible options for using this model for the purpose of a scientifically grounded approach to the development of these technologies in the country are discussed.

Land is split in to land parcels supported size and ownership. This land parcel is that the lowest unit in local level planning of villages in Indian context. During this research, a study was conducted to reinforce land parcel level Biocapacity using soil and water as parameters at villages of Gudiyattam Block, Vellore District. The Gudiyattam block consists of 44 panchayat villages including Gudiyattam town spatially encompasses an outsized area. Planning for such a large region requires significant resources of manpower, time and technology. Additionally, the enhancement of Biocapacity requires land-parcel level information. With the given time and resources, it's impossible to do that. Therefore, Gudiyattam block has been divided into performance zones. From each performance zone one panchayat village is taken into account for the analysis. This research used available data for a given year and validated the land parcel level planning as a reliable method for enhancement of biocapacity. This method also has the potential to be used with high-resolution satellite imagery to capture crop data periodically and use that as a basis for computing biocapacity at land parcel level. Land parcel level planning enables precision farming with high yields to reinforce biocapacity. This research used current crop data at land parcel level to spot the potential for higher yields to reinforce Biocapacity.

Herbicide use is rising globally to enhance food production, causing harm to environment and the ecosystem. Precision agriculture suggests variable-rate herbicide application based on weed densities to mitigate adverse effects of herbicides. Accurate weed density estimation using advanced computer vision techniques like deep learning requires large labelled agriculture data. Labelling large agriculture data at pixel level is a time-consuming and tedious job. In this paper, a methodology is developed to accelerate manual labelling of pixels using a two-step procedure. In the first step, the background and foreground are segmented using maximum likelihood classification, and in the second step, the weed pixels are manually labelled. Such labelled data is used to train semantic segmentation models, which classify crop and background pixels as one class, and all other vegetation as the second class. This paper evaluates the proposed methodology on high-resolution colour images of canola fields and makes performance comparison of deep learning meta-architectures like SegNet and UNET and encoder blocks like VGG16 and ResNet-50. ResNet-50 based SegNet model has shown the best results with mean intersection over union value of 0.8288 and frequency weighted intersection over union value of 0.9869.

In this paper we consider the multiple line detection problem by using the center-based clustering approach, and propose a new incremental method based on iterative reweighted approach. We prove the convergence theorem and construct an appropriate algorithm which we test on numerous artificial data sets. A stopping criterion in the algorithm is defined by using the parameters from the DBSCAN algorithm. We give necessary conditions for the most appropriate partition, which have been used during elimination of unacceptable center-lines that appear in the output of the algorithm. The algorithm is also illustrated on a real-world image coming from Precision Agriculture.

In this paper we assess the sustainability of Variable Rate Technology (VRT) application for different wheat intensive agricultural areas in Europe. Within the H2020 FATIMA project we developed an analytical framework to assess the environmental and economic sustainability of nutrient management in two Case Studies in Greek and Czech Republic, implemented with different fertilization strategies in the winter wheat production, compared with the conventional one. We look at the economic and environmental dimension expressed throughout a set of specific performance indicators implemented under the Water Energy Food (WEF) nexus framework. Considering that fertilization strategies significatively affect environmental impact as well as economic performance at farm level, we investigated how Precision Agriculture (PA) tools, and particularly VRT, most suitable for increasing efficiency in fertilization distribution, can improve and maximize efficiency of inputs and profitability of individual fields by targeting application where needed and at optimum rates. The study shows how, despite a general reduction of production cost and increase in gross margin, farm sizes and the level of efficiency of the “as usual scenario” influence the economic impact of the VRT. Looking instead at the environmental perspective of the analysis, the results provide evidence of a decrease of nitrogen (N) due to lower amount of fertilizers, as well as a better distribution according to the potential productivity of the field, causing reduced impact of groundwater quality (nitrate loss by leaching). On the other side, some constraints arise from the high cost and the investments needed to adopt VRT at farm level, one of the main challenges to be solved through adequate policy instruments.

In this research a machine learning framework was developed for cotton yield estimation using multi-temporal remote sensing data collected from unmanned aircraft system (UAS). The proposed machine learning model was based on an artificial neural network (ANN) and used three types of crop features derived from UAS data to predict the yield, namely; multi-temporal features including canopy cover, canopy height, canopy volume, normalized difference vegetation index (NDVI), excessive greenness index (ExG); non-temporal features including cotton boll count, boll size and boll volume, and irrigation status as a qualitative feature. The model provided low residual values with predicted yield values close to the observed yield values (R2 ~ 0.9). ANN model performance was compared with support vector regression (SVR) and random forest regression (RFR). Comparison results revealed that ANN model outperforms SVR and RFR. Additionally, redundant features were removed using correlation analysis, and an optimal subset of features was obtained that included canopy volume, ExG, boll count, boll volume and irrigation status. Moreover, the relative significance of each feature in the optimal input feature subset was determined using sensitivity analysis. It was found that canopy volume and ExG contributed around 50% towards the corresponding yield. Finally, further analysis was performed to investigate how early in the growing season the model can accurately predict yield. It was observed that even at 70 days after planting the model predicted yield with reasonable accuracy (R2 of 0.72 over test set). This study revealed that UAS derived multi-temporal data along with non-temporal and qualitative data can be combined within a machine learning framework to provide a reliable estimation of crop yield and provide effective understanding for crop management.

Thermographies are a source of rich information, valuable in precision agriculture tasks such as crop stress assessment, plant disease analysis, and soil moisture evaluation. Traditionally, practitioners obtain soil temperature from the ground or using satellites and other airborne methods, which are costly and offer limited spatial and temporal resolution. In this paper, we introduce a method to measure soil surface temperature dynamics with the use of an unmanned aerial system (UAS). In our approach, we fuse information from thermal and multispectral cameras with ambient variables to generate estimates for soil temperature using computational intelligence models. Using the images, we produce a spatial reconstruction using structure from motion (SfM). After the multimodal registration of the resulting geo-referenced orthomosaics, we characterize the dynamics of the soil surface temperature using the differences between consecutively captured temperature orthomosaics. In our results, we are capable of estimating soil surface temperature from a UAS flying at 30 m AGL with a RMSE of 3.24∘C ± 0.3 and 1.77∘C ± 0.2, at one standard deviation, for two test fields with average ground sampling distances below 6.0 cm/pixel, using a Random Forest regressor.

Stomata is an important channel for the exchange of water and gas between crops and the external environment. Stomatal behavior can shed light on how environmental stressors affect crop growth. Traditional stomata investigation methods involve experimental laboratory measurements of specific plants using stoma-related instruments or a field plot. However, developing a new remote sensing canopy resistance model for analyzing the stomatal behavior of crops under typical environmental stress is essential for regional agricultural management. Canopy resistance is a vital microscale–macroscale (from kilometers to micrometers) bridge that represents the stomatal behavior of the entire crop. In this study, a remote sensing canopy resistance model was proposed based on the Penman-Monteith model and a remote sensing evapotranspiration model. The model was verified by field-observed winter wheat data at the Yucheng Comprehensive Experiment Station (YCES) with dry-hot wind (DHW) and brackish water irrigation treatments from 2017 to 2019. The model measured canopy resistance (regression coefficient = 1.25; R2 = 0.98) well. Finally, remote sensing-based canopy resistance measurements were used to investigate the effects of soil salinity and dry-hot wind on the winter wheat. The results showed excellent model performance for retrieving crop stomatal behavior under two environmental stresses and proved that the proposed remote sensing canopy resistance model is a promising tool for the investigation of stomatal behavior under environmental stress; this is particularly relevant for regional precision farming.

Continuous monitoring of the diameter and rings of woody plants is required to evaluate agroforestry products such as timber. Manual measurement is still widely used to monitor tree growth but it is time-consuming. In addition, advanced equipment and software are available but they require technical skills and are costly for smallholders. However, we have developed a rapid and low-cost sensing system for estimating the diameters of woody plants. The apparatus includes a sensor and Internet of things (IoT) technology, which can be used under any network conditions. The apparatus was tested with various trees and the results were compared with those obtained by manual measurement. The findings using the newly developed system demonstrate a strong correlation (R2 = 0.99) with the manually performed measurements. This rapid and low-cost diameter sensing system will be useful in the expansion of agroforestry sensing networks and beneficial to smallholders. Hence, the system will assist in informing foresters of potential problems in timely manner.

Accurate and automatic vegetation extraction from digital plant images in the field is a widely studied topic in precision agriculture. Many techniques focus on pixels or regions to be segmented as plants or back-grounds, such as colour index-based and learning-based methods. Different from a traditional two-class classification problem, the proposed method regarded vegetation extraction as a multi-class task. In consideration of manually annotated errors at the edge of a plant image, the original marked mask was re-labelled using a Gaussian probability function. To capture more adequate information in the process of feature extraction, 9 pixel-level colour features and 18 region-level statistical characteristics of neighbourhood pixels were computed from three colour spaces. The extracted 27-dimensional features were inputs of a classification model, which output multi-class labels. A suitable threshold was finally selected to obtain the segmented image. Experimental results showed that the proposed multi-class and multi-level features (MCMLF) method achieved better performance than the other approaches. Through the quantitative and qualitative analysis of segmentation results, it was also found that the suggested method had high computation efficiency as well as strong adaptation ability to solve the outdoor challenges, including various lighting conditions, shadow regions, and complex backgrounds.

Remote sensing technology has long been used to detect and map crop diseases. Airborne and satellite imagery acquired during growing seasons can be used not only for early detection and within-season management of some crop diseases, but also for the control of recurring diseases in future seasons. With variable rate technology in precision agriculture, site-specific fungicide application can be made to infested areas if the disease is stable, although traditional uniform application is more appropriate for diseases that can spread rapidly across the field. This article provides a brief overview of remote sensing and precision agriculture technologies that have been used for crop disease detection and management. Specifically, the article illustrates how airborne and satellite imagery and variable rate technology have been used for detecting and mapping cotton root rot, a destructive soilborne fungal disease, in cotton fields and how site-specific fungicide application has been implemented using prescription maps derived from the imagery for effective control of the disease. The overview and methodologies presented in this article should provide researchers, extension personnel, growers, crop consultants, and farm equipment and chemical dealers with practical guidelines for remote sensing detection and effective management of some crop diseases.

Quantifying canopy biophysical parameters is critical to agricultural research and farm management. In this study, strawberry dry biomass and leaf area were modeled statistically using high spatial and temporal resolution imagery. A mobile field data acquisition system was used to acquire thousands of very high resolution (~0.5 mm) close-range images seven times throughout the strawberry growing season. Ortho-mosaics and dense point clouds were generated through Structure from Motion (SfM) and used in Object-Based Image Analysis (OBIA) at the sub-leaf level to extract canopy structure variables such as planimetric canopy area, canopy average height, and canopy smoothness metric. Regression analysis was carried out using these image-derived canopy variables as predictors to model leaf area ( R 2 = 0.79; ten-fold cross-validation RMSE = 0.056 m2) and dry biomass ( R 2 = 0.84; ten-fold cross-validation RMSE = 7.72 g) obtained through destructive measurements. Results indicate consistent predictive power through the season and across 17 strawberry genotypes. The study showed that the canopy smoothness metric developed in this study as an indicator of canopy density could complement other variables (planimetric canopy area, canopy average height) that describe canopy geometric properties.

To analyse the adoption of drones in agriculture as the latest tool added to the set of precision agriculture technologies, this paper makes use of a novel application of the trans-theoretical model of behavioural change by analysing a sample of 167 large-scale German farmers collected in 2019. The model provides a gradual measure of farmers' decision making with respect to the adoption of drones, which gives more detailed insight into farmers' adoption processes than the more common approach of applied binary classifications. Ordinal logit regression results show that, among other factors, farmers' age, precision agriculture technology literacy and farm size affect farmers’ adoption process. Thus, this paper contributes to the literature by identifying key determinants of the drone adoption process in agriculture. Furthermore, this study provides information about the fields of drone application as well as reasons that oppose the usage of drones by farmers. The results are of interest for policy makers and suppliers of drones.

Recent research studies and policies about innovative solutions to reduce water and energy consumption in food production are briefly reviewed. Options to increase water use efficiency and productivity include soil mulching, drip irrigation, deficit irrigation, and precision agriculture. As for the energy–water nexus, attention is focused on energy audits of water distribution networks; improving of system performance –– network sectoring, use of variable speed drives, critical points control, electricity tariff — and reduction of wastewater treatment’s energy use. At a larger scale, other solutions emerge: diversification and rotation of crops, cultivation of drought-resistant crops, and optimization process of the spatial distribution of cropping patterns. The rebound effect that can be associated to these options is also considered.

Site-specific application of herbicides is highly desirable for optimising its usage and reducing environmental damages. Thus, developing techniques for identification and mapping of weeds is necessary for a proper precision agriculture adoption. Such weed identification for site-specific management is difficult when the main crop is already established in the field. This study shows the possibility of differentiating sugarcane plants from weeds by the spectral behaviour of the leaves. The performance of two modelling methods, SIMCA (soft independent modelling by class analogy) and the RF algorithm (random forest) was compared. The simplification of the Vis-NIR spectrum into only four bands of interest (500–550 nm; 650–750 nm; 1300–1450 nm; and 1800–1900 nm) was verified by demonstrating they had the same differentiation ability as the full visible-near infra-red spectrum. Thus, it was shown that performing the proper band selection and local calibration using a spectral classification approach may allow weed mapping and facilitate localised herbicide application.

The massive growth of wireless communications in recent years is mostly due to new connectivity demands and advances in technology development of low power) transceivers. An example of the unique demands is the increasing exchange of data in Internet services, which has led to wireless network deployment for data transmissions. The coordination of the IoT devices, smart systems, and agriculture can contribute directly to the development of the farmer’s practices by building their farm more intelligent and digital. However, enhancing farming practices requires inspecting farm equipment and farmer’s experiences, which can be analyzed through the interconnectedness of IoT objects to collect farm data over the Internet to launch smart digital agriculture. It is challenging to control all farming processes (especially in real-time), this remaining as the main limitation of traditional farming. In this work, we focus on how wireless sensors can play a vital role in smart farm systems and allow processing the large amount of data generated in batches or real-time to analyze it, retrieve insights from it, and create a Smart Digital Farm. This paper proposes hierarchical-logic mapping and deployment algorithms to tackle the problem of poor network connectivity and sensing coverage in random IoT deployment.

This study aims to examine the problem of sampling interval selection for the precision agriculture using Wireless Sensor Networks (WSNs). The Wireless Sensor Networks is introduced for limited energy sensor nodes and deployed in an agricultural farm, which is divided into a number of heterogeneous agricultural areas. A Variable Sampling Interval Precision Agriculture (VSI-PA) system is proposed to measure and monitor the agricultural parameters for appropriate agricultural activities e.g. water irrigation. At every time step, the sensor nodes are selected for each area to sense the soil moisture and temperature. After that, a variable sampling interval is calculated independently for each individual area, according to the soil temperature of the area. Analysis of simulation results show that, compared with other fixed sampling interval schemes, the proposed VSI-PA system provides a significant improvement in energy consumption while maintaining a small variation of the soil moisture regardless of the values of soil temperature. Therefore, the crop yields have been enhanced. Keywords: Variable sampling interval, Energy-efficient, Irrigation system, Precision agriculture, Wireless sensor networks

The world’s population growth in the last decades demanded a great increase in agricultural production, especially in food. Therefore, in order to reduce possible losses and guarantee productivity, farmers depend more and more on the application of agrochemicals in their crops. This massive use of pesticides represents not only a high cost to the farmers but also a risk to their health, to the environment and even to the safety of the food consumed by the population. In this context, new technologies have been developed to make agricultural spraying more effective, reducing the amount of pesticide applied and dosing its use according to the need of the crop. The recognition of its spatial and temporal variability is treated by Precision Agriculture. In the case of pest management it can be done using variable-rate sprayers or using on/off application with individual nozzle control. Thus, this paper proposes a low-cost prototype in a modular solution to automate existent agricultural sprayers. The solution allows individual nozzle opening, with on/off control, using solenoid valves, pressure and flow sensors, Arduino boards, and smartphone. Additionally, the prototype has a data logger function to store nozzle status and sensor values, allowing future analysis and application reports.

Increasing environmental concerns are a driving force in the search for ways to improve the efficiency of mineral nitrogen (N) fertilization. Spectral sensors to determine the crop’s supply status are among the most mature precision agriculture technologies to adapt the N dose site-specifically to the crop’s need. By using artificial intelligence techniques like expert systems based on fuzzy set theory, the algorithms of such sensor systems could be adapted by the farmer to the highly varying conditions among the specific fertilization dates. This paper is dealing with the development of a fuzzy logic based model of the commercial Yara N-Sensor’s dosing algorithms. Simulations for several sets of input-output data acquired in field experiments showed high accordance with the behaviour of the N-Sensor system with good adaptability to different calibrations.

Unmanned Aerial Vehicles or UAVs can be employed in a multitude of civil applications owing to their ease of use, low maintenance, affordability, high-mobility, and ability to hover. Such vehicles are being utilized for real-time monitoring of road traffic, providing wireless coverage, remote sensing, search and rescue operations, delivery of goods, security and surveillance, precision agriculture, and civil infrastructure inspection. They are the next big revolution in technology and civil infrastructure is expected to dominate their more than $45 Billion worth market. This paper surveys the UAV assisted Structural Health Monitoring or SHM literature over the last decade and categorizes UAVs based on their aerodynamics. Further, it presents the payload product line to facilitate the SHM tasks, details the different applications of UAVs exploited in the last decade to support civil structures and discusses the key challenges faced in its application across various domains.

Grape is non-climacteric fruit. For export quality grapes, farmers regularly monitor plants for identification of possible diseases and spray relevant pesticides. Due to time constraints and effort involved, farmers usually spray an excessive pesticide which causes health hazards. Mechatronics system is proposed to monitor health of the grapes leaves based on histogram analysis using the MATLAB. It alert farmers about the initiation of major grape diseases viz. Downey Mildew, Powdery Mildew and Anthracnose and trigger automatic spraying of pesticides based on the philosophy of Precision Farming, results economical utilization of pesticides and save their valuable time for other relevant agricultural activities.

Agriculture plays an important role in the life span of human being not only for their survival but for the better economic growth of the country too. Precision agriculture is the new trending term in the field of technology whose main motive is to reduce the workload of the farmers and increase the productivity of the farms by using technologies like I.O.T, WSNs, Remote Sensing, Drone surveillance and many more. In this paper, we show the work done by our cost effective and reliable device whose aim is to irrigate fields only when there is a need of water and to provide information about detection of any intrusion in agricultural fields. The information is sent to the farmers by using cloud application. The performance of our system is measured in terms of intrusion detection and moisture of soil for irrigation.

Agricultural herbicide application efficiency can be improved using smart sprayers which provide site-specific, rather than broadcast, applications of agrochemicals. The YOLOv3-Tiny convolutional neural network (CNN) was trained to detect two weeds, hair fescue and sheep sorrel, in images captured from wild blueberry fields throughout Nova Scotia, Canada. An evaluation was performed in three commercial wild blueberry fields in Nova Scotia to examine the effects of camera selection and target distance on detection accuracy. A Canon T6 DSLR camera, an LG G6 smartphone, and a Logitech c920 webcam were used to capture RGB images at varying distances from target weeds. Mean F1-scores for each combination of camera and image height were analysed in a 3 × 3 factorial arrangement for hair fescue and a 3 × 2 factorial arrangement for sheep sorrel. Images captured from 0.98 m with the LG G6 and Canon T6 produced F1-scores of up to 0.97 for detection of at least one hair fescue tuft. Images captured with the LG G6 and Canon T6 DSLR from 0.57 m achieved F1-scores of 0.94 and 0.93, respectively, for detection of at least one sheep sorrel plant per image. Sheep sorrel was undetectable in images from the Logitech c920 under 19 of 27 parameter combinations. Future work will involve using the CNN to control herbicide applications with a real-time smart sprayer. Additionally, the CNN will be used in a web-based application to detect target weeds and provide site-specific information to aid management decisions. Using a CNN to detect weeds will create improvements in management techniques, resulting in cost-savings and greater sustainability for the wild blueberry industry.

Improvements in the sustainability of agricultural production depend essentially on advances in the efficient use of nitrogen. Precision farming promises solutions in this respect. Variable rate technologies allow the right quantities of fertilizer to be applied at the right place. This helps to both maintain yields and avoid nitrogen losses. However, these technologies are still not widely adopted, especially in small-scale farming systems. Recent developments in sensing technologies, like drones or satellites, open up new opportunities for variable rate technologies. In this paper, we develop a bio-economic modelling framework to assess the usefulness of different sensing approaches in variable rate fertilization to measure environmental heterogeneity at field level, ranging from satellite imagery to drones and handheld N-sensors. We assess the utility of these sensing technologies and quantify the effects on yields, nitrogen input and associated net returns using wheat production in Switzerland as our case study. Our results show that net profits increase when a high-resolution technology is applied to fields which exhibit higher spatial heterogeneity of soil conditions and lower spatial autocorrelation of different soil types. However, even with a high degree of spatial heterogeneity within a field, both the overall utility of variable rate fertilization and the absolute differences in the net returns between the technologies remain low. Our results suggest that the additional cost of using a drone that provides the highest resolution should not exceed 4.5 CHF/ha compared to the use of a standard N-sensor or satellite imagery. Thus, the adoption of variable rate technologies depends essentially on the additional economic and environmental effects they generate. Therefore, it might be necessary to implement specific policy measures, such as taxes on nitrogen in combination with subsidies. Moreover, specific technology providers, such as contractors, may play a vital role in technology uptake since the economic benefits might only play out at larger spatial levels., Ecological Economics, 185, ISSN:0921-8009

Humanity is exceeding the planetary boundary for nitrogen. Changing course will depend, in part, on how current and future smallholders manage nitrogen. Precision agriculture could play a key role but only if accompanied by policy reforms and investments that address a range of barriers to improved nitrogen management.

Characterizing the spatial variability of soil thermal properties and quantifying their relationships with commonly measured soil physical properties is important for successful implementation of precision agriculture practices. However, such information at field scale is scarce. This study aimed to characterize the spatial variability of soil thermal properties and to quantify their spatial relationships with soil physical properties at a field scale. Measurements of soil water content (SWC) and thermal properties were carried out at 80 locations using a time domain reflectometry (TDR) probe and a heat pulse probe (Decagon KD2 Pro), respectively in a cultivated cornfield in southwestern QC, Canada. Soil texture and total organic carbon (TOC) at those locations were also determined in the laboratory. Both coefficient of variation (CV%) and variogram range showed a weak to moderate variability and spatial dependency for the soil properties which was attributed to the homogeneity of the soil, crop, and land management practices. Results of principle component analysis (PCA) indicated that the soil thermal properties and soil water content (SWC) had strong correlations with principle component 2 (PC2). The variogram ranges indicated that the soil thermal properties varied on a similar spatial scale as the SWC in the field. Cross variograms indicated strong spatial correlations between SWC and thermal properties, particularly thermal conductivity (k) and thermal diffusivity (D). Overall, results of this study showed that the SWC had a stronger influence on the spatial variability of soil thermal properties than the other properties such as TOC and soil texture at the field scale. Therefore, the results of this study could be useful for accurate modeling of the soil thermal regimes and mass-energy exchange processes at the field scale as well as developing successful agricultural management strategies such as precision irrigation. Future studies should explore the relationships between the thermal properties and other soil properties such as bulk density under various cropping systems.

Fusarium damage in wheat reduces the quality and safety of associated food and feed products. In this study, a specific Fusarium head blight (FHB) classification index (FCI) for detection of this disease in wheat is proposed. Hyperspectral microscopy images of wheat spikelets are used as the data source. An algorithm combining the instability index (ISI) and spectral angle mapper (SAM) classifier (ISI-SAM) is used to extract four sensitive single wavelengths. Then, partial least-squares regression of the disease area ratio with simple spectral vegetation indices (SVIs) for each image is used to determine the most relevant spectral index (i.e., difference spectral index, DSI (668, 417)). On this basis, in order to develop a hyperspectral index for FHB detection, an exhaustive search for the best weighted combination of a single wavelength and DSI (668, 417) is conducted, with all possible combinations being tested. The final FCI is F C I = 0.25 × [ 2 ( R 668 − R 417 ) − R 539 ] , with an overall classification accuracy of 89.80%. The FCI is tested for its ability to detect and classify the healthy and diseased areas of wheat spikelets through comparison with six commonly used SVIs, and its disease identification accuracy is almost 30% higher than that of the best-performing SVI. The FCI is also successfully applied to classification of hyperspectral image data from wheat spikes. The FCI developed in this study constitutes a stable and feasible method of early FHB detection through spectral and low-altitude remote sensing, and will improve FHB detection, identification, and monitoring performance in precision agriculture applications.

Stress in crops is one of the major concerns in precision agriculture because it indicates the emergence of disease and damage in plants. Detecting the stress condition of a plant early is critical. A system that can monitor the condition of plants is a desirable solution. In this work, Collaborative Control Theory is utilized to construct a new system, ARS (agricultural robotic system) which synchronizes humans, a mobile robot, and a variable set of sensors to effectively perform the monitoring and detection tasks. A key protocol for that system, which combines routing algorithm, adaptive search algorithm, and collaboration control framework has been developed and validated, and is presented in this article. By using greenhouse as a case study structure, the protocol routes a robot to visit the sampled locations by using a genetic algorithm. In addition, the search algorithm can be guided by the predictive characteristics of the crops’ stress, which can spread to other plants according to sunlight, airflow direction, and other known conditions. Based on simulation experiments, the results indicate with statistical significance that (1) the routing algorithm increases the number of successful detections of existing stressed plants by 45.77% compared to monitoring without this routing algorithm. (2) The adaptive search algorithm improves the number of successful detections of stressed plants by 71.88% compared to a system without the adaptive search algorithm. (3) The new protocol developed in this research yields the highest overall robotic efficiency, compared with a system without collaborative control framework.

New and improved drought models based on the World Meteorological Organization approved Standardized Precipitation Index, principally at multiple timescale horizons, are providing significant benefits to the hydrological community, by its widespread acceptance in the sub-field of water resources management, sustainable water use and precision agriculture. In this research paper, the existing challenges faced by a drought forecasting model trained at multiple time-scales are resolved where a new multi-phase, multivariate empirical mode decomposition model integrated with simulated annealing and Kernel ridge regression algorithms (i.e., MEMD-SA-KRR) is designed to attain significantly accurate drought forecasts for 3 agricultural sites (i.e., Faisalabad, Islamabad and Jhelum, located in Pakistan). Utilizing the multi-scalar Standardized Precipitation Index (SPI) time series as a target variable for characterization of drought, twelve multivariate datasets (derived from statistically significant lagged combinations of precipitation, temperature & humidity), that are enriched with eight synoptic-scale climate mode indices and periodicity, are utilized in designing a new drought model. The study constructs a hybrid MEMD-SA-KRR model, where firstly, the data are partitioned into their respective training and testing subsets after creating historically lagged SPI at timescale (t – 1). Secondly, the MEMD algorithm is conditioned to demarcate multivariate climate indices from their training and testing sets, separately, into their decomposed intrinsic mode functions (IMFs) and residues. Thirdly, the SA method is employed to decide the most suitable IMFs. Finally, the KRR algorithm is applied to the selected IMFs to forecast multi-scaler SPI, at 1-, 3-, 6- and 12-monthly forecast horizon. The results are benchmarked with Random Forest, integrated with MEMD and SA to develop the MEMD-SA-RF equivalent model. The multi-phase MEMD-SA-KRR model is tested geographically in Pakistan, revealing that the MEMD-SA-KRR hybrid model generates reliable performance in forecasting multi-scaler SPI series, relative to the comparative models based on error analysis metrics. The hybrid drought model incorporating the most pertinent synoptic-scale climate drivers, as the model inputs has significant implications for hydrological applications and water resources management including its potential use in drought policy and drought recovery plans.