Corn Canopy Research Articles

Numerical simulation of agricultural unmanned helicopter corn canopy spraying

Plant protection by unmanned aerial vehicle (UAV) is a highly complicated process, especially considering the vast and diversified crop canopy. Furthermore, the droplet dispersal mechanism above and inside the crop canopy remains unclear. Owing to this, the authors used the lattice Boltzmann method (LBM) to simulate the application of plant protection UAV. The LBM is suitable not only to simulate the rotor tip vortex of UAV but also to model the crop canopy with detailed geometry. It is found that the streamwise and spanwise speed of the downwash flow achieve an extreme value both on top of the corn canopy and near the ground. When the droplet diameter decreased from 400 to 50 μm, its deposit region width doubled and its floating time increased from 3 s to more than 12 s. Small droplets are more likely to be lifted by the secondary flow near the ground and drift for a longer distance, while larger droplets are deposited directly. The increase in application height weakened the secondary spanwise flow near the ground and caused less droplet re-lifting and spanwise movement. The best point to initiate and terminate spraying is 2.5 m after passing the target corn canopy region. The droplet penetration rate in the canopy decreased with an increase in droplet size. This study used the LBM to model the unmanned helicopter application on a full corn canopy. Droplet deposition and distribution correspond to droplet size and canopy structure. The simulation results establish a foundation to optimize agricultural aerial spraying in an efficient and methodical manner.

Dynamic changes in evapotranspiration, canopy photosynthesis and expansive growth in open field under rapid fluctuating radiation

In open fields, plants experience dynamic changes in environment, particularly radiation, temperature, wind, and humidity but their short-term responses have not been adequately characterized under natural conditions. In this study, we assessed causal effects of rapid radiation fluctuation on seven plant parameters in open fields of cotton and sweet corn canopies. The parameters are evapotranspiration (ET), canopy photosynthesis indicated by CO2 flux from air to canopy (FCO2), sensible heat energy flux (H), canopy conductance for water vapor (gcw), canopy surface temperature (Ts), shoot and leaf elongation rate, and stem diameter change. The energy and CO2 fluxes were measured with Bowen ratio/energy balance/CO2 gradient (BREB+) technique, using averaging time of 5 min. Shoot + leaf elongation and stem diameter change were monitored with position transducers using averaging time of 1 min. All parameters were all found to respond to change in radiation and transpiration within minutes or sooner. While radiation effects on canopy gas exchanges are expected, illuminating are the indirect but immediate effects on shoot + leaf growth and stem diameter change through radiation effects on transpiration and plant water status. A novel finding is that gcw also responded within minutes or sooner to radiation fluctuations and that FCO2 was related almost linearly to gcw. Results are discussed in terms of soil-plant-atmosphere continuum, and interpreted in terms of dynamic interactions between transpiration and plant water status. The clear inverse relationship between ET and elongation rate or stem diameter changes provides additional evidence supporting the validity of 5-min averaging for the BREB + technique.

A corn canopy organs detection method based on improved DBi-YOLOv8 network

Corn canopy organs detection is critical in obtaining high-throughput phenotypic data. Accurate identification of each organ can provide a reliable data source for canopy phenotype determination, which has significant theoretical and practical value for corn variety breeding, cultivation management, and high-quality and high-yielding production. Due to the difficulty in quickly identifying corn canopy organs in the natural environment of the field, it is challenging to obtain high-throughput phenotypic data. Therefore, this paper proposed a method for corn canopy organs detection based on an improved network model (DBi-YOLOv8). Firstly, the Raspberry Pi 4B was used as the sensor control center to construct an embedded system for corn canopy image acquisition and collected 987 images of corn plants. Secondly, the improved deformable convolution and Bi-level routing attention were embedded into the backbone and neck structures of the YOLOv8 network. With training the improved network, a corn canopy detection model was obtained, which enabled the rapid detection of corn canopy organs. Finally, the LTNS algorithm and TBC algorithm were proposed for counting of the number of leaves, ears, and tassels. On the testing set data, the detection performance of the model was analyzed through different evaluation metrics. The results showed that the mAP and FPS of the detection model were 89.4% and 65.3, which increased by 12% and 0.6 compared to the original model. In addition, both algorithms have high reliability, with the coefficient of determination R2 for counting crown leaves, ears, and tassel branches being 0.9336, 0.8149, and 0.917, respectively. This achievement proposed an accurate, non-destructive, and fast corn canopy organs detection model, providing reliable technical support for quantifying various traits of corn plants, field crop growth monitoring, and elite variety breeding.

The Throughfall, Stemflow, and Canopy Interception Loss in Corn and Soybean Fields in Northeast China

Information about throughfall, stemflow, and canopy interception loss is essential for the water use efficiency of crops and the dynamic processes of water erosion. A two-year field experiment was conducted under natural rainfall conditions to observe the characteristics and factors that affect throughfall, stemflow, and canopy interception loss in corn (Zea mays L.) and soybean (Glycine max L.) fields in northeast China from 2019 to 2020. Nine measurement sites (A, B, C, D, E, F, G, H, and I) were distributed horizontally between two planting rows under the crop canopy. The mean value of the throughfall volume (TF) in measurement locations B, C, and G under the corn canopy and measurement locations B and C could represent the mean level of TF of corn and soybean fields, respectively. The volume of TF, stemflow (SF), and canopy interception loss (CI) of two growing seasons from 2019 to 2020 accounted for approximately 58.5%, 30.1%, and 11.4% of the gross rainfall (GR) of two growing seasons in corn fields, and 78.0%, 7.5%, and 14.5% of the GR in the soybean field, respectively. The TF and TFR of each rainfall event in the corn and soybean fields could be predicted by linear regression models with a normalized root mean square error (NRMSE) lower than 25.0%. These results and prediction models will be used in water management and soil erosion control in northeast China.

A solution to the problem of height overcompensation in GNSS-IR corn height measurements: a canopy reflectance calibration model

ABSTRACT The global navigation satellite system interferometric reflectometry (GNSS-IR) technique for collecting corn height has the limitation of not accounting for changes in the corn canopy reflector. A model that uses the corn canopy reflection coefficient to account for changes in the canopy’s reflector and the power attenuation of the signal reflected by the soil surface is proposed. This letter proposes a method to solve the overcompensation problem of traditional GNSS-IR vegetation height measurement. Considering the decrease in GNSS signal penetration depth and the upward shift of the actual reflecting surface caused by changes in corn leaf water content (LWC), this letter provides a calibration model for the GNSS-IR corn height measurement. The GNSS signals reflected from the corn canopy were separated using empirical mode decomposition (EMD), and the height measurement was then calibrated using the directed and reflected signal ratios. The results of the experiments show that the model proposed in this letter can effectively improve the accuracy of corn height measurement, and the root mean square error (RMSE) is reduced to 0.17 m . Compared with the current study, the model proposed in this study reduces the influence of the actual reflective surface variation of the corn canopy on the height measurement using GNSS-IR.

UAS and UGV-Based Disease Management System for Diagnosing Corn Diseases Above and Below the Canopy Using Deep Learning

Highlights Generated custom imagery dataset using UAS, UGV, and handheld sensors separately in diseased corn fields. Proposed data pipeline utilizing Google Sheets API to establish communication between each platform and enable access through a web application. Developed deep learning-based disease management system for above and below-the-canopy corn disease diagnosis. Trained and evaluated disease detection models from each platform separately to provide management recommendations. Abstract. Early disease management following the onset of disease symptoms is crucial for controlling their spread. Heterogenous collaboration between unmanned aerial systems (UAS) and unmanned ground vehicles (UGV) for field scouting and disease diagnosis is a potential approach for developing automated disease management solutions. However, automation of crop-specific disease identification requires the use of above and below-canopy sensors and properly trained deep learning (DL) models. This research proposes to develop a novel disease management system for diagnosing corn diseases from above and below the canopy by collaboratively using edge devices mounted on UAS and UGV, respectively. Three separate datasets were acquired using UAS above the canopy, UGV below the canopy, and handheld imaging platforms within diseased corn fields. DL-based image classification models were first trained for identifying common corn diseases under field conditions, resulting in up to 95.04% testing accuracy using the DenseNet169 architecture. After creating bounding box annotations for disease images, You Only Look Once (YOLO)v7 DL-based object detection models were trained to identify diseases from each platform separately. Trained YOLOv7 models resulted in the highest mAP@IoU=0.5 of 37.6%, 46.4%, and 72.2% for locating and identifying diseases above the canopy using UAS, below the canopy using UGV, and handheld sensors, respectively. A client/server architecture was developed to establish communication between the UAS, UGV, and Google Spreadsheets via Wi-Fi communication protocol. The coordinates of diseased regions and distinct disease types were recorded using the client/server architecture on Google Spreadsheets. A web application utilized the data from the Google Spreadsheet to help users diagnose diseases in real time and provide recommendations for implementing appropriate disease management practices. This study reports findings of independently operated UAS and UGV that can potentially offer disease spread from below and over the corn canopy by combining the information. Keywords: Deep learning, Disease identification, Disease management, Object detection, UAS, UGV, YOLOv7.

Light partitioning strategies impact relative fitness of weeds and cover crops when drill-interseeding in corn

AbstractDrill-interseeding cover crops into corn (Zea mays L.) is an emerging establishment method in northern U.S. production regions. However, cover crop performance in interseeded systems remains variable, and creating environments that are conducive to cover crop but not weed growth is challenging. Cultural practices that partition resources between corn and interseeded cover crops have potential to improve performance if weeds are adequately managed. This study evaluated interactions among corn hybrids differing in leaf architecture (upright, pendulum), corn row spacing (76 cm, 152 cm), and interseeding timing (V3, V6) on light transmittance, relative fitness of cover crop species (cereal rye [Secale cereale L.], annual ryegrass (Lolium multiflorum Lam), red clover [Trifolium pratense L.]) and weeds, and corn grain yield at three U.S. Northeast locations. Results showed that light transmittance through the corn canopy was greater in 152-cm row spacing compared with 76-cm row spacing at the V6 growth stage, with the magnitude of difference increasing at the V10 corn growth stage. Corn hybrids had a marginal effect on light transmittance. The effect of row spacing and interseeding timing on fall cover crop biomass varied across cover crop species and locations. In 76-cm rows, interseeding earlier (V3) increased cover crop biomass production. The relative fitness of cover crops was greater than that of weeds in each combination of cultural practices that included narrow spacing (76 cm), whereas the relative fitness of weeds was greater than that of cover crops when interseeding in wide rows (152 cm). The effect of row spacing on corn yield varied among locations, with higher yields observed in 76-cm row spacing compared with 152-cm at two of three locations. Our results show that interseeding early (V3) on 76-cm row spacing can balance cover crop and corn production management goals, while placing cover crops at a relative fitness advantage over weeds.

The roles of radiative, structural and physiological information of sun-induced chlorophyll fluorescence in predicting gross primary production of a corn crop at various temporal scales

Extensive research suggests that sun-induced chlorophyll fluorescence (SIF) and gross primary productivity (GPP) have a near-linear relationship, providing a promising avenue for estimating the carbon uptake of ecosystems. However, the factors influencing the relationship are not yet clear. This study examines the roles of SIF's radiative, structural, and physiological information in predicting GPP, based on four years of field observations of a corn canopy at various temporal scales. We quantified SIF's radiative component by measuring the intensity of incident photosynthetically active radiation (iPAR), and separated the structural and physiological components from SIF observations using the fluorescence correction vegetation index (FCVI). Our results show that the R2 values between SIF and GPP, as estimated by linear models, increased from 0.66 at a half-hour resolution to 0.86 at a one-month resolution. In comparison, the product of FCVI and iPAR, representing the non-physiological information of SIF, performed consistently well in predicting GPP with R2>0.84 at various temporal scales, suggesting a limited contribution of the physiological information of SIF for GPP estimation.. The results further reveal that SIF's radiative and structural components positively impacted the SIF-GPP linearity, while the physiological component had a negative impact on the linearity for most cases, changing from 0.6 % to -27.5 %. As for the temporal dependency, the controls of the SIF-GPP relationship moved from radiation at diurnal scales to structure at the seasonal scales. The structural contribution changed from 14.8 % at a half-hour resolution to 92.4 % at a one-month resolution, while the radiative contribution decreased from 118.0 % at a half-hour resolution to 11.7 % at a one-month resolution. This study contributes to enhancing our understanding of the physiological information conveyed by SIF and the factors influencing the temporal dependency of the SIF-GPP relationship.

Visible and Near-Infrared Hyperspectral Diurnal Variation Calibration for Corn Phenotyping Using Remote Sensing

Remote sensing coupled with hyperspectral technology has become increasingly popular to investigate plant traits, showcasing its advantages in studying plant growth, health, and productivity. The quality of the collected hyperspectral images is crucial for subsequent data analysis and plant phenotyping studies. However, diurnal variations in spectral characteristics introduce more data variance in canopy reflectance spectra, raising the cost of subsequent analyses and compromising the performance of trait estimation models. In this study, a fixed gantry platform in a cornfield was used to capture visible and near-infrared (VNIR) hyperspectral images of corn canopies at consecutive time intervals. By applying reference board calibration and locally weighted scatterplot smoothing to minimize the effects of ambient light and daily growth, diurnal spectral changes across all involved VNIR wavelengths were investigated. Several distinct diurnal patterns were observed to have close connections with the plants’ physiological effects. Diurnal calibration models were established at every wavelength by employing the least squares polynomial algorithm, with the highest coefficient of determination reaching 0.84. Moreover, by employing diurnal calibration in canopy spectra processing, the reduction in spectral variance brought about by varying imaging time was evidently exhibited. This study not only reveals the diurnal spectral variation pattern at VNIR bands but also offers a reliable, straightforward, and low-cost approach to improve the quality of remote sensing data and reduce the inherent variance brought about via the different imaging times ensuring that comparable spectral analysis can be performed under relatively fair conditions.

Spatial corn canopy temperature extraction: How focal length and sUAS flying altitude influence thermal infrared sensing accuracy

Irrigation water management starts with quantifying irrigation prescriptions based on crop water requirements at a spatial scale. The most accepted and effective method for estimating this water need is to use the canopy temperature of the plants. The use of small unmanned aerial systems (sUAS) with a thermal infrared camera has long been established as an effective method of measuring plant canopy temperatures at a spatial scale. However, concerns still exist about the accuracy of canopy temperatures collected by these systems and how imagery collection altitude or camera viewing angle affects the temperature estimation. To address these concerns, this study was designed to evaluate the effects of flying altitude of airborne thermal infrared (TIR) imagery and camera view angle on corn-canopy temperature sensing accuracy and image quality. Three different thermal cameras with focal lengths of 9 mm, 13 mm, and 19 mm were used, and each camera was flown at an altitude of 30 m, 50 m, and 70 m. Thermal cameras were mounted on an sUAS and flown over a 6000 m2 cornfield. The sUAS was flown at a speed of 3 m/s on autopilot mode guided by pre-planned missions. The thermal camera was triggered once per second, and imagery was geotagged using an external GPS unit. Images obtained using different combinations of focal lengths and flight altitudes were processed and converted into orthomosaics. Calibration was conducted using ground-based temperature reference panels and temperature maps were created from the orthomosaics. The orthomosaics were examined for the accuracy of the corn-canopy temperature sensing, ability to differentiate between hot and cold surfaces, ease of image stitching/developing orthomosaics, geometric accuracy, image quality, and spatial resolution. The results indicated that with the combination of appropriate camera focal length, altitude, and image calibration techniques, a canopy temperature map of crops with a temperature error of less than 2 °C from the actual canopy temperature can be produced. A narrow-angle thermal camera flying at low altitudes (<50 m) was found to be the least suitable combination for corn canopy temperature sensing. The most appropriate combination for temperature estimation of corn canopies was with a 13 mm focal length camera flying at an altitude of 50 m above ground level.

Simultaneous Retrieval of Corn Growth Status and Soil Water Content Based on One GNSS Antenna

The retrieval of crop growth status using Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) has become a major area of interest within the field of vegetation remote sensing in recent years. Using only a single GNSS antenna, it is difficult to determine the crop growth status and soil water content (SWC) in vegetation-covered regions due to plenty of multi-path signals. Based on the empirical mode decomposition and the spectrum difference, this study presents an algorithm that can decompose and separate signals reflected by the soil surface or corn canopy. Because the low-roughness soil surface is isotropic while the corn canopy is anisotropic, the signals reflected by the soil surface have a higher proportion of coherent components than those reflected by the corn canopy. The moduli between the retrieved heights and the actual heights (for the same interval from different satellites) have the least variance. In this study, the signals reflected by the soil surface and the corn canopy are separated using the variance of retrieved heights. When the corn grows taller than the GNSS antenna, the vegetation water content (VWC) of the corn leaves becomes the primary factor affecting the direct signal’s intensity, as the leaves obstruct the signal. Hence, the VWC of corn leaves can be calculated through the power attenuation of signals. An experiment performed on a plot of land covered with corn shows that, after multi-GPS-satellite fusion, the correlations between the retrieved corn canopy height, leaf VWC, soil water content (SWC), and in situ data reach 0.94, 0.92, and 0.88, respectively. The corresponding root mean square errors are 0.195 m, 0.0055 kg/cm2, and 0.0484 cm3/cm3, respectively.

Evaluation of the Tau-Omega Model Over a Dense Corn Canopy at P- and L-Band

Evaluation of the Tau-Omega Model Over a Dense Corn Canopy at P- and L-Band

Growth and development of multiple waterhemp (Amaranthus tuberculatus) cohorts in corn and soybeans

Waterhemp is a troublesome weed species in cropping systems throughout Wisconsin and much of the US Midwest. Coupled with extended emergence window, rapid growth, high genetic diversity, and herbicide resistance to multiple sites of action, waterhemp represents a major challenge for chemical-based weed control. The objective of this experiment was to evaluate the impact of soybean and corn canopy on growth and development of multiple waterhemp cohorts in 2019 and 2020. Treatments consisted of narrow- (38 cm row spacing) and wide-row (76 cm row spacing) soybeans, corn (76 cm row spacing), and fallow (no crop), with 6 waterhemp transplant timings (cohorts) in 2019 and 2020, respectively. Waterhemp seedlings grown in the greenhouse were transplanted at the 2-3 true leaves growth stage to the field at 10-day increments throughout the typical waterhemp emergence season for Wisconsin (June-July). Waterhemp plants were measured for height and harvested for biomass when they reached the flowering stage. The number of days from the first transplant timing to result in 50% and 90% biomass and height reduction (ED50 and ED90) of subsequent waterhemp cohorts were estimated and used as indicators of weed suppression ability by the respective crop. Narrow-row soybeans required fewer days to suppress 50% growth of new waterhemp cohorts followed by corn, wide row soybeans, and fallow, respectively (4, 6, 9, and 14 d for 50% biomass reduction, and 10, 14, 18, and 42 d for 50% height reduction, respectively). Similarly, narrow-row soybeans required fewer days to suppress 90% biomass of waterhemp, followed by corn, wide-row soybeans, and fallow (11, 15, 18, and 78 d, respectively). Conversely, narrow and wide-row soybeans required fewer days to suppress 90% height of new waterhemp cohorts followed by corn and fallow (20, 26, 43, and 85 d, respectively). Rapid soybean canopy closure (e.g., planting soybeans earlier and in narrow row-spacing) combined with other integrated weed management practices have the potential to reduce reliance on POST herbicide application and mitigate herbicide resistance evolution. In corn, mid- to late-season monitoring and management of later emerging waterhemp cohorts is recommended due to longer time required to reach 90% waterhemp growth suppression.

Towards Understanding the Influence of Vertical Water Distribution on Radar Backscatter from Vegetation Using a Multi-Layer Water Cloud Model

For a good interpretation of radar backscatter sensitivity to vegetation water dynamics, we need to know which parts of the vegetation layer control that backscatter. However, backscatter sensitivity to different depths in the canopy is poorly understood. This is partly caused by a lack of observational data to describe the vertical moisture distribution. In this study, we aimed to understand the sensitivity of L-band backscatter to water at different heights in a corn canopy. We studied changes in the contribution of different vertical layers to total backscatter throughout the season and during the day. Using detailed field measurements, we first determined the vertical distribution of moisture in the plants, and its seasonal and sub-daily variation. Then, these measurements were used to define different sublayers in a multi-layer water cloud model (WCM). To calibrate and validate the WCM, we used hyper-temporal tower-based polarimetric L-band scatterometer data. WCM simulations showed a shift in dominant scattering from the lowest 50 cm to 50–100 cm during the season in all polarizations, mainly due to leaf and ear growth and corresponding scattering and attenuation. Dew and rainfall interception raised sensitivity to upper parts of the canopy and lowered sensitivity to lower parts. The methodology and results presented in this study demonstrate the importance of the vertical moisture distribution on scattering from vegetation. These insights are essential to avoid misinterpretation and spurious artefacts during retrieval of soil moisture and vegetation parameters.

Downwash characteristics and analysis from a six-rotor unmanned aerial vehicle configured for plant protection.

One theoretical advantage of using unmanned aerial vehicles (UAVs) to spray pesticides for maturing corn is that the strong downwash penetrates canopies. However, only few studies have been conducted to examine in-canopy downwash characteristics. This paper investigated the downwash by a six-rotor UAV in mature cornfields. 3D wind speeds in corn canopies and an open area were measured, and comparisons conducted. The downwash by the UAV resulted in in-canopy maximum wind speeds. Z-dimensional downwash was sensitive to all factors, whereas the X- and Y-dimensional downwashes were related to layers and crop positions. Meanwhile, when comparing with the downwash between a 2m hovering position and the optimal flight parameters, the X-dimensional and Y-dimensional motion time of top-layer downwash generally advanced by 3.8s and 1.6s, whereas both motion time and the strength of the Z-dimensional downwash were impeded by ≈2.2-s hysteresis at middle layers and ≈4.5-s time reduction, respectively. Thus, combined with distributions, the corn on the left or right might not be sprayed sufficiently. Furthermore, under the convergence requirement error of 0.01, the overall correlation of the model was ≈0.846 in terms of the Z-dimensional downwash and ≈0.55 and 0.61 for the X- and Y-dimensions, respectively. The selection of operation parameters should mainly consider the Z-dimensional downwash. The optimal operation parameters were a height of 2m with a speed of 4m s-1 . Meanwhile, the canopy effect could influence the uniformity, motion and strength of downwash. Predictions could be achieved before operation. © 2022 Society of Chemical Industry.

Corn Nitrogen Nutrition Index Prediction Improved by Integrating Genetic, Environmental, and Management Factors with Active Canopy Sensing Using Machine Learning

Accurate nitrogen (N) diagnosis early in the growing season across diverse soil, weather, and management conditions is challenging. Strategies using multi-source data are hypothesized to perform significantly better than approaches using crop sensing information alone. The objective of this study was to evaluate, across diverse environments, the potential for integrating genetic (e.g., comparative relative maturity and growing degree units to key developmental growth stages), environmental (e.g., soil and weather), and management (e.g., seeding rate, irrigation, previous crop, and preplant N rate) information with active canopy sensor data for improved corn N nutrition index (NNI) prediction using machine learning methods. Thirteen site-year corn (Zea mays L.) N rate experiments involving eight N treatments conducted in four US Midwest states in 2015 and 2016 were used for this study. A proximal RapidSCAN CS-45 active canopy sensor was used to collect corn canopy reflectance data around the V9 developmental growth stage. The utility of vegetation indices and ancillary data for predicting corn aboveground biomass, plant N concentration, plant N uptake, and NNI was evaluated using singular variable regression and machine learning methods. The results indicated that when the genetic, environmental, and management data were used together with the active canopy sensor data, corn N status indicators could be more reliably predicted either using support vector regression (R2 = 0.74–0.90 for prediction) or random forest regression models (R2 = 0.84–0.93 for prediction), as compared with using the best-performing single vegetation index or using a normalized difference vegetation index (NDVI) and normalized difference red edge (NDRE) together (R2 &lt; 0.30). The N diagnostic accuracy based on the NNI was 87% using the data fusion approach with random forest regression (kappa statistic = 0.75), which was better than the result of a support vector regression model using the same inputs. The NDRE index was consistently ranked as the most important variable for predicting all the four corn N status indicators, followed by the preplant N rate. It is concluded that incorporating genetic, environmental, and management information with canopy sensing data can significantly improve in-season corn N status prediction and diagnosis across diverse soil and weather conditions.

Spatial Corn Canopy Temperature Extraction: How Focal Length and Suas Flying Altitude Influence Thermal Infrared Sensing Accuracy

Spatial Corn Canopy Temperature Extraction: How Focal Length and Suas Flying Altitude Influence Thermal Infrared Sensing Accuracy

Post‐harvest drone flights to measure weed growth and yield associations

AbstractDrone flights are often only performed during the growing season, with no data collected once harvest has been completed, although they could be used to measure winter annual weed growth. Using a drone mounted with a multispectral sensor, we flew small plot corn (Zea mays L.) fertility, cover crop, and population studies at black layer and 0–14 d after harvest (DAH). Yields had positive correlations to normalized difference vegetation index (NDVI) at black layer but often had negative correlations to corn yields 0–14 DAH. After harvest, NDVI could be associated with weed growth, and negative correlations to yield could point to reduced corn canopy allowing light to reach late‐season weeds. In fertility studies, excess nitrogen appears to increase weed biomass after harvest, which can be easily identified through drone imagery. Flights should be performed after corn harvest as weed growth may provide additional insight into management decisions.

Ground versus aerial canopy reflectance of corn: Red‐edge and non‐red edge vegetation indices

AbstractConstant‐rate blanket applications of fertilizer N can result in both an over and under supply relative to crop utilization on a field‐by‐field basis. Variable‐rate (VR) applications tailored to better meet crop demand can improve N use efficiency on spatially variable soils. The objectives of this study were to compare the response in corn (Zea mays L.) canopy reflectance derived vegetation indices (VI) to varying fertilizer N rates and to determine relationships between resulting VIs acquired using two different sensing platforms. Four fertilizer N rates in 50/50 split at V1–2 and V6–7 leaf stages were applied, from deficient to excessive, to create varying corn nutritional N status. Sensing and biophysical sampling were performed throughout the season for analysis and comparison to calculated VIs. Grain yield plateaued around 135±10 kg N ha–1 across the study. Furthermore, strong relationships between VIs and fertilizer N rates were found, with the strongest using combined indices that incorporate the red‐edge wavelength (720 nm). Relationships strengthened at later growth stages. The response models were found to be sensor specific, VI specific, and mostly non‐transferable between sensors. Results from this study demonstrate the utility of using remote sensing technologies to predict corn N status more accurately for eventual use in VR prescription development.

Estimation of corn yield based on hyperspectral imagery and convolutional neural network

Corn is an important food crop in the world, widely distributed in many countries because of its excellent environmental adaptability. Moreover, corn is an important feed source for animal production and it is an indispensable raw material for many different industries. With increasing human population and decreasing arable land, there is an increased focus on increasing yield of corn.Convolutional neural network (CNN) analysis can be used for non-destructive yield prediction and is well suited for classification and feature extraction. The overall objective of this experiment was to use hyperspectral imagery to train a CNN classification model to estimate corn grain yield. High resolution hyperspectral imagery was captured at five corn growth stages - V5 (five leaves with visible leaf collars), V8 (eight leaves with visible leaf collars), V10 (ten leaves with visible leaf collars), V12 (12 leaves with visible leaf collars), and R2 (blister stage). Hyperspectral imagery was denoised using the wavelet analysis method, then was used to train and validate the CNN model. The spectral information reflecting the internal characteristics and the spatial information provided by the color image (red, green and blue bands extracted from hyperspectral image) reflecting the external characteristics of corn growth are extracted for modelling and verification. The results show that the spectral and color image-based integrated CNN model has a classification accuracy of 75.50%. In contrast, the accuracy of a one-dimensional CNN model based only on spectral information or a two-dimensional CNN model based only on color image information were 60.39% and 32.17%, respectively. The integrated CNN model (spectral information plus color image information) is better than results of the individual one-dimensional CNN or two-dimensional CNN models. In addition. The Kappa coefficient of integrated CNN model is 0.69, which indicates a high consistency of classification. Comprehensive use of spectral information and color image information, which represent information about the inner and outer corn canopy can provide more accurate corn yield prediction than one-dimensional or two-dimensional CNN models.