introduction
Precision agriculture (PA) has developed intensively over the last decade in response to the growing demand for food for the growing population. PA technologies ensure the efficient use and management of available resources to increase crop production, reduce freshwater consumption, maintain soil fertility, and protect the environment. Unlike conventional agricultural technologies, PA uses a variable application approach for cultivation practices (Shafi et al., 2019).
The effective implementation of AP is highly dependent on the availability and cost of advanced tools and technologies suitable for the agricultural industry, such as smart machinery and robotics, global positioning system (GPS), sensor networks, and monitoring systems, acquisition and information processing, etc. One of the key components of PA is a remote sensing/monitoring instrumentation responsible for measuring and processing data on plant and soil conditions. Remote monitoring works in conjunction with GPS to provide a georeferenced map of the characteristics of the field used for cultivation (Tsouros et al., 2019, Maes and Steppe, 2019, Kashyap and Kumar, 2021).
In the field of remote sensing, soil moisture monitoring is of particular interest due to global climate change in key agricultural regions and the extremely high proportion of fresh water used for irrigation. Campbell et al., 2017, reported that more than 70% of the world's fresh water is used for agricultural purposes. Soil water content primarily determines the biophysical processes that affect crop production and soil health. Therefore, a comprehensive and detailed knowledge of soil moisture is an important input parameter for a public address system to control water application and consumption. Further processing of georeferenced moisture data by advanced digitized farm management brings significant improvement in freshwater conservation and increase in crop production (Kashyap and Kumar, 2021).
In recent decades, various soil moisture measurement technologies based on different physical principles have been developed and are applicable to PAs. These technologies differ in terms of objective purpose, practical application, precision, cost, weight, etc. However, the most suitable soil moisture measurement solution for PA is a highly accurate, lightweight, mobile instrumentation system that allows for long/close distance measurement. -range measures. The demand for such a system is related to the growing group of drone-based surveillance applications integrated with public address infrastructure. Mobile sensor equipment is also suitable for integration into smart agricultural machines and robotics (Tsouros et al., 2019, Maes and Steppe, 2019, Inoue, 2020).
Traditional methods for determining soil moisture are generally based on stationary/well sensors or soil samples taken for analysis in a laboratory. These methods use a variety of physical principles, eg. B. Gamma ray laboratory analysis of a soil layer sample; the neutron scattering technique in wells (neutron probe); various electromagnetic sensor technologies, etc. (Kodikara et al., 2014, Balaghi et al., 2018, Babaeian et al., 2019) Conventional methods are actually point measurement techniques, making the development of a soil moisture map extremely difficult when a number of measurement points is not enough. Obviously, these methods cannot be used in applications where the mobility of the detection devices is essential.
Several advanced methods have been developed that provide remote measurement of soil moisture for aerial and satellite monitoring purposes. Soil moisture remote measurement technologies are generally represented by two methods: active measurement, where the instrument reads the reflected electromagnetic energy (radar), and passive measurement, where the energy emitted by the earth is recorded (radiometry). Remote sensing systems used in satellite Earth observation are designed for large-scale operation or for global mapping. Two projects have been launched to collect soil moisture data through satellite monitoring systems: (1) NASA's Soil Moisture Active Passive (SMAP) mission (Entekhabi et al., 2010, Chan et al., 2016) and (2) Soil Moisture and Ocean Salinity (SMOS) from the European Space Agency (ESA) (Kerr et al., 2001, Mecklenburg et al., 2012). However, the measurement accuracy of satellite monitoring systems is not suitable for AP and smart agriculture. The best precision of SMAP is 3 km in radar mode (Chan et al., 2016), while the spatial precision of the SMOS system is around 40 km (Mecklenburg et al., 2012).
Proximal methods developed in recent years include various ground-based radiometers and radar systems, as well as imaging and electromagnetic induction techniques. Proximal systems for soil moisture measurement are considered the best candidates for mobile/drone public address applications. The most promising technique is microwave radiometer technology. It is a passive microwave remote sensing instrument that measures the parameters of natural radiation emitted from the earth. The recorded emission data represents the brightness temperature of the Earth's surface (Ulaby and Long, 2014). Post-processing of the data converts it into the soil moisture value, which is linked to a specific geographic location. Microwave radiometer technology offers the opportunity to develop a soil moisture monitoring instrument with reduced size, weight, and cost.
The dielectric constant of the soil depends on many parameters. However, the dominant parameter that affects the dielectric constant is the biochemical composition of the soil. If the biochemical composition is defined, the main factor influencing the dielectric constant of the soil is the moisture content. Vegetation cover does not influence the dielectric constant of the soil, but it can lead to inaccuracies in the measurement process. This means that the best measurement accuracy can be obtained when the field has no vegetation cover or when the size of the vegetation is small.
This article proposes a new approach for a software algorithm for a microwave radiometer-based system to measure the soil moisture content of the observed earth. The proposed approach ensures rapid processing of the acquisition data to provide the surface dielectric permittivity, which is associated with the soil moisture content. The paper also discusses the details of design, development, and practical implementation of a low-cost, lightweight mobile instrumentation device for measuring and processing soil moisture data using the proposed approach. The instrument is designed to work in conjunction with GPS technologies and integrate with a public address management system. It is designed for use in drones, portable and mobile applications.
section cutouts
principle of operation
Microwave radiometry as a remote sensing method goes back to the work of Jansky and Dicke on the study of relict radiation (Tomiyasu, 1974). Originally, radio astronomers used ultrasensitive receivers of radio emissions from various objects and environments, called microwave radiometers. Subsequently, in the 1960s and 1970s, microwave radiometers for spacecraft installations were successfully developed and operated to provide observation and surveillance from space.
Proposed algorithmic approach
According to the proposed approach, two 3D temperature surfaces (Figs. 3 and 4) are intersected by a horizontal plane corresponding to a given radio brightness temperature measured by the radiometer in horizontal and vertical polarization. The cross section method produces two 2D curves in the plane with two axes: (x axis) real and (y axis) imaginary parts of the observed dielectric constant of the surface. Combined curves in the same plane intersect at a
Instrument prototypes
A prototype of the dual-polarization L-band portable radiometer was designed and built to measure soil temperature and moisture content using the proposed approach. The radiometer prototype was tested through a series of practical experiments to verify the performance of the software algorithm described in Fig. 6. Figure 7 shows a view of the prototype device. The main parameters of the radiometer are given in Table 1.
The radiometer circuit is based on the well-known diagram of
Field test results and discussion
For a practical field test, the microwave radiometer was mounted on a cart as shown in Fig. 12 and transported across the test field along pre-planned line tacks. Vertically and horizontally polarized radio brightness values were recorded in the logger's memory simultaneously with data from the GPS navigation system and internal sensors. Two reference sources were calibrated at the end of each shift. Relic radiation was used as a "cold" source,
conclusions
The document presents the development and construction details of the low-cost, lightweight mobile microwave radiometer prototype for measuring and processing soil moisture data. The instrument works according to the proposed software algorithm, which allows fast processing of the acquisition data. Using Earth radio brightness temperature readings, the software algorithm calculates the complex dielectric permittivity of the surface, which is related to the moisture content of the soil. The instrument
Conflict of Interest Statement
The authors declare that they are not aware of any competing financial interests or personal relationships that may have influenced the work described in this paper.
gracias
The research was funded by the Russian Scientific Foundation (Project No. 19-19-00349) and by a grant from the Federal State Budgetary Institution "Fund for Supporting the Development of Forms of Small Businesses in the Scientific and Technical Field" (Fund of Support for Innovation) contract n° 2886ГC1/ 45415. This document was supported by the Strategic Academic Leadership Program of the RUDN University.
Featured Articles (6)
Investigation article
Use spectral indices and soil attribute data sets and combine them to predict cadmium levels in agricultural soils
Computers and Electronics in Agriculture, Volume 198, 2022, Article 107077
The continued demand for arable land to produce an optimal crop depends on the continuous application of agrochemicals and fertilizers to increase soil fertility and control disease. The successive application of fertilizers and the use of agrochemicals associated with the metallurgical and steel industry introduce potentially toxic elements into the soil. Active agricultural activities and industrial emissions leading to the injection of cadmium (Cd) into the atmosphere and active deposition in agricultural soils (particularly from the primary metals industry, steel and iron industries). The cadmium concentration in the study area exceeds the local background value. As a result, the excessive concentration of cadmium in the soil contributes to increased toxic and carcinogenic effects, with negative impacts on the environment and human health. Therefore, determining the spatial distribution of Cd is crucial for environmentally friendly agricultural production and the reduction of Cd emissions in soils. The objectives of this study are (i) to determine the variability of Cd prediction in agricultural soils using spectral indices or terrain attributes together with modeling algorithms and (ii) to determine if the combination of spectral indices and terrain attributes together with algorithms Modeling Foresight can improve efficiency on agricultural soils. The study applied three modeling scenarios: prediction using terrain attributes together with digital soil mapping (DSM) approaches (scenario 1), prediction using spectral indices in combination with DSM (scenario 2), and prediction using a combination of terrain attributes. , spectral indices and DSM (Scenario 3). Gaussian Process Regression (GPR), Partial Least Squares Regression (PLSR), Extreme Gradient Boosting (EGB), Multivariate Adaptive Regression Splines (MARS), Bayesian Regularized Neural Network (BRNN), Regularized Random Forest (RRF), Model Bayesian Generalized Linear (BGLM) and M5 tree models were the DSMs used in the study. The M5 tree model and terrain attributes {Scenario 1 R2= 0.77, Correlation Coefficient of Concordance (CCC) = 0.73, Root Mean Square Error (RMSE) = 0.45, Mean Absolute Error (MAE) = 0.37, and Mean Absolute Error (MoAE) = 0.35 }, GBS and spectral indices {Scenario 2, R2=0.83, CCC=0.76, RMSE=0.54, MAE=0.33 and MoAE=0.23} and the M5 tree model, spectral indices and terrain attributes {Scenario 3, R2=0.84, CCC=0.81, RMSE=0.39, MAE=0.31 and MoAE=0.24} were the best overall combinations to predict Cd in agricultural soil. The overall evaluation of the approaches suggested that the combination of spectral indices, terrain attributes, and the M5 tree model in Scenario 3 was the optimal technique for predicting Cd in agricultural soils. Therefore, a combination of environmental covariates with a high correlation with the response variable combined with appropriate modeling techniques that predict potentially toxic elements in agricultural soils will give the best results.
Investigation article
Mobile detection to estimate hydrodynamic parameters for minimally studied open channels
Computers and Electronics in Agriculture, Volume 198, 2022, Article 107072
To ensure an equitable distribution of surface waters to open water channels, it is important to monitor these channels with high spatial and temporal resolution. However, channel monitoring is a challenging task when channels are minimally inspected, i. h have limited water level and water flow sensors to describe channel dynamics. Such sensors are usually installed in control structures (gates) along a channel. When control structures are far apart, these sensors cannot accurately describe water levels and velocities in a channel with high spatial resolution. Furthermore, it is not economically feasible to install these sensors on all channels, as they require permanent installation and infrastructure costs. Therefore, inspired by low-cost mobile sensing and its portability aspect, this paper offers the formulation of a framework that takes into account GPS measurements of mobile sensing to provide channel dynamics with higher resolution. We demonstrate the effectiveness of the proposed framework in comparison to fixed sensor frames using an experimental data set obtained from an irrigation canal in Lahore, Pakistan. The proposed framework uses the State Dependent Interactive Multiple Models (SD-IMM) technique and the estimation results show improvements compared to the results obtained from fixed sensors along the channel.
Investigation article
Pig surveillance scheme based on identification and counting by integrating BLE tags and WBLCX antennas
Computers and Electronics in Agriculture, Volume 198, 2022, Article 107070
Recently, Internet of Things (IoT) technologies have been applied to monitoring pigs. Among other things, electronic tags help identify individual fattening pigs. However, tags are not enough to achieve accurate and reliable monitoring on pig farms. Instead, advanced processing data is required for individual behaviors, such as the times of visits to feeding and resting areas and the number of movements between these areas. Numerical measurement of individual pig movements in the pen is a prerequisite for obtaining these data. This article proposes a surveillance scheme for a large herd of pigs for individual identification and counting. As a technical solution, a surveillance system is being developed using Bluetooth Low Energy (BLE) tags and Wireless Broadband Leaky Coaxial Cable (WBLCX) antennas. The monitoring system collects credentials transmitted by the BLE tags attached to individual pigs. The monitoring scheme then estimates the location and movement of each pig by calculating the collected data. A series of experiments were performed to evaluate the detection properties between BLE tags and WBLCX antennas and demonstrate their performance. In addition, the effectiveness of the proposed area determination algorithm and monitoring system was verified through field tests at a pig farm in Japan. Our experimental results are summarized as follows. First, data comparisons between the area determination algorithm and visual inspections confirmed the validity of the proposed surveillance scheme. Second, area movements and residence times of pigs in a pigsty could be examined by the monitoring scheme. Based on the experimental results, it is expected that with the use of the proposed surveillance, the locations and movements of large herds of pigs in the pigsty can be observed. This article proposes a surveillance scheme based on the identification and counting of individual pigs, explains the details of the surveillance scheme based on hardware configurations, and verifies its effectiveness and feasibility through experiments.
Investigation article
Automatic detection method for trailer hopper for silage harvesting based on improved U-Net
Computers and Electronics in Agriculture, Volume 198, 2022, Article 107046
At present, automatic silage filling of self-propelled silage harvesters is mainly carried out by computer vision; Tipper detection accuracy is one of the main factors affecting the reliability of the launch arm control. However, due to the limitations of complicated harvesting environment, variable lighting and different types of trailers, traditional trailer hopper detection methods cannot meet the operational needs of silage harvester with high efficiency and low loss. In this article, we propose a new semantic segmentation network called RSHC U-Net to achieve a better trailer bin segmentation effect on the silage harvest image. The network was developed on the basis of U-Net, which effectively includes the Residual Dissolution (RC) module, the Expression and Excitation Attention Mechanism (SE) module, the Hybrid Exempt Dissolution (HDC) module, and the Cross Integrated Level. Collection Decoder Module (CLGD). The RC module solves the problem of decreasing slope with increasing depth of the network; the SE module highlights the characteristics of the edge of the funnel; the HDC module compensates for the loss of detailed information in the encoder; the CLGD module bridges the semantic gap between encoder and decoder. The test results show that the average MIoU, Acc, Pre, and ST of the RSHC U-Net are 90.38%, 95.17%, 90.22%, and 300.4 ms, respectively. The interior point ratio threshold for the trailer body tuning algorithm is set to 95%. The elapsed time, accuracy of fit, and mean absolute error of the algorithm are 91.43 ms, 94.75%, and 9.71, respectively. Finally, the spatial coordinates of the trailer hopper vertex are retrieved from the ZED API to perform spatial positioning of the hopper, providing a basis for subsequent planning of the silage filling route and control of arm movement. launch. This project has a certain technical reference value to fill silage quickly, accurately and automatically.
Investigation article
Low Range Constraint-Based Imaging Algorithm for Detecting Straw Sieve Clogging in Corn Harvesters
Computers and Electronics in Agriculture, Volume 198, 2022, Article 107056
Although a debris screen is used to separate contaminants from the kernels during corn harvest, it often becomes clogged with contaminants, negatively affecting harvesting efficiency. Accurate image recognition is the most important step in automatically adjusting working parameters to avoid clogging. Unfortunately, sieve meshes between contaminants cannot be detected using existing algorithms, and the masked area of the mesh and the background contaminant cannot be easily distinguished. To address this problem, a screen clogging detection (LSCR) algorithm based on low range constraints is proposed in this study. Unlike existing algorithms, the position and shape of the meshes are accurately estimated using the low-rank optimization strategy and no training patterns or comprehensive information related to the mesh contour of the target images is needed. The congestion areas are then determined based on the difference in relative reflectivity. The experimental results show that the overall detection accuracy at the pixel level using the LSCR algorithm reaches 0.943 for the test scenes, which is significantly higher than that of existing algorithms. LSCR can potentially be used for in-line detection of chaff screen clogging in corn combines.
Investigation article
How Internet use affects the agricultural land rental market: an empirical study from rural China
Computers and Electronics in Agriculture, Volume 198, 2022, Article 107075
The agricultural land rental market plays a crucial role in China. This study investigates the impact of the Internet on the leasing of agricultural land by rural households. The study uses data from 2018 China Family Panel Studies (CFPS) surveys and a binary logit model to examine the influence of Internet use from four aspects, including Internet use, type, time, and frequency of Internet use. Internet grouped for different purposes. The results show that homeowners who use the Internet are more likely than non-users to rent farmland. The more intensive use of the Internet (more hours) is directly correlated with the use of the Internet in the agricultural land leasing market. The use of the Internet through the mobile phone has a greater influence on the leasing of agricultural land than the use through the computer. A more detailed analysis shows that Internet use significantly improves the social networks of rural households and increases their likelihood of engaging in non-agricultural business and employment activities, thus contributing to agricultural land tenure decisions. The government should improve the information infrastructure and provide information management training in rural areas to encourage leasing of agricultural land.
© 2022 Elsevier B.V. All rights reserved.