Sources of angle-dependent errors in terrestrial laser scanner-based crop stand measurement

In past years, research was conducted to investigate the potential of laser rangefinders and scanners for vehicle-based measuring of crop stand physical parameters. High coefficients of determination (R² > 0.90) were found between the crop biomass and the laser rangefinder and scanner readings in the form of the mean height of reflection points of the laser beam. It was observed that the height of the reflection point increased depending on the detecting angle of the laser beam in a considerable manner. This phenomenon indicates that farther crop plants generate higher reflection points, resulting in an overestimation of crop height respectively of crop biomass. In the face of these unsolved problems, the object of the paper was to investigate the measuring properties of a chosen laser scanner depending on the inclination angle and the scanning angle and to analyze the error sources for vehicle-based laser scanner measurements in crop stands. Therefore, the scanner was investigated in two test series (May 30, 2008 in winter wheat, and June 10, 2009 in winter rye) along a transect (tramline) with a length of approximately 700 m. The performed comparisons demonstrated that one part of the observed overestimation of the reflection point height can be explained theoretically by the geometry of the laser beam. The main part of overestimation was explained by the recognizability of the gap fraction in crop stands. Because no sufficient theoretical compensation algorithm for overestimation resulting from gaps exists, it must be concluded that for each specific laser rangefinder type and crop species, the overestimation depending on detection angles has to be investigated individually in field tests.     

Laser rangefinder-based measuring of crop biomass under field conditions

Knowledge of site-specific crop parameters such as plant height, coverage and biomass density is important for optimising crop management and harvesting processes. Sensors for measuring crop parameters are essential pre-requisites to gather this information. In recent years, laser rangefinder sensors have been adopted in many industrial applications. In agricultural engineering, the potential of laser rangefinders for measuring crop parameters has been little exploited. This paper reports the design and the performance of a measuring system based on a triangulation and a time-of-flight laser rangefinder for estimating crop biomass density in representative crops under field conditions. It was shown that the mean height of reflection point is a suitable parameter for non-contact indirect measurement of crop biomass by laser rangefinder sensors. The main parameters for potential assessment were the coefficient of determination (R ² ) and the standard error (RMSE) for the relation between crop biomass density and the mean height of the reflection point in crop stands from oilseed rape, winter rye, winter wheat and grassland during the vegetation period in 2006. For the triangulation sensor, R ² was in the range from 0.87 to 0.98 and for the time-of-flight sensor in the range from 0.75 to 0.99 for both fresh matter and dry matter density. The triangulation sensor had a reduced suitability caused by masking effects of the reflected beam and because of limited measuring range. Based on the results of experiments and technical data, it was concluded that the time-of-flight principle has good potential for site-specific crop management.     

Crop height variability detection in a single field by multi-temporal terrestrial laser scanning

Information on crop height, crop growth and biomass distribution is important for crop management and environmental modelling. For the determination of these parameters, terrestrial laser scanning in combination with real-time kinematic GPS (RTK–GPS) measurements was conducted in a multi-temporal approach in two consecutive years within a single field. Therefore, a time-of-flight laser scanner was mounted on a tripod. For georeferencing of the point clouds, all eight to nine positions of the laser scanner and several reflective targets were measured by RTK–GPS. The surveys were carried out three to four times during the growing periods of 2008 (sugar-beet) and 2009 (mainly winter barley). Crop surface models were established for every survey date with a horizontal resolution of 1 m, which can be used to derive maps of plant height and plant growth. The detected crop heights were consistent with observations from panoramic images and manual measurements (R² = 0.53, RMSE = 0.1 m). Topographic and soil parameters were used for statistical analysis of the detected variability of crop height and significant correlations were found. Regression analysis (R² < 0.31) emphasized the uncertainty of basic relations between the selected parameters and crop height variability within one field. Likewise, these patterns compared with the normalized difference vegetation index (NDVI) derived from satellite imagery show only minor significant correlations (r < 0.44).     

Evaluation of the use of LIDAR laser scanner to map pruning wood in vineyards and its potential for management zones delineation

Vine vigour assessment has been a major concern of precision viticulture studies in order to identify areas of uniform vine performance within vineyards. Moreover, the counting and weighing of winter dormant canes is considered as the most informative measurement to indicate vine balance and is commonly performed manually by grape growers for management purposes. The main concern of this measurement is that it is time consuming and laborious and it cannot accommodate detailed sampling density. In the present study, the potential of using laser scanner technology as an automated, easy and rapid way to perform mapping of the winter pruning wood across the vineyard was investigated. The study was conducted during 2010 and 2011, in a one hectare commercial vineyard in central Greece, planted with cv. Agiorgitiko, a traditional Greek variety for the production of red wine. Parameters of topography, soil depth, soil texture, canopy properties (NDVI), yield, and grape quality were mapped and analysed in conjunction to winter canes weighing at pruning time. The mapping of the dormant canes was carried out using a 2D laser scanner sensor prior to pruning and manually measuring the pruning weight on a 10 × 20 m grid. Laser scanner measurements showed significant relationship in both 2010 and 2011 with pruning weight (r = 0.809 and r = 0.829 respectively, p < 0.001), yield and early season NDVI, showing the potential of using laser scanner measurements to assess variability in vine vigour within vineyards. These results suggest that laser scanners offer great promise to characterize within field variability in vine performance.     

Methodology for measuring fAPAR in crops using a combination of active optical and linear irradiance sensors: a case study in Triticale (X Triticosecale Wittmack)

The amount of photosynthetically active radiation (PAR, 0.4–0.7 μm) absorbed by plants for photosynthesis relative to incident radiation is defined as the fraction of absorbed photosynthetically active radiation (fAPAR). This is an important variable in both plant biomass production and plant growth modeling. This study investigates the application of a newly developed, linear irradiance sensor (LightScout Quantum Bar Sensor, LightScout, Spectrum Technologies, Inc. USA), to quantify fAPAR for a demonstrator crop, Triticale (X Triticosecale Wittmack). A protocol was devised for sensor placement to determine reflected PAR components of fAPAR and to determine the optimal time of day and sensor orientation for data collection. Coincident, top of canopy, normalized difference vegetation index (NDVI) measurements were also acquired with a CropCircle? ACS-210 sensor and measurements correlated with derived fAPAR values. The optimum height of the linear irradiance sensor above soil or plant canopy was found to be 0.4 m while measuring reflected PAR. Measurement of fAPAR was found to be stable when conducted within 1 h of local solar noon in order to avoid significant bidirectional effects resulting from diurnal changes of leaf orientation relative to the vertically-placed sensor. In the row crop studied, averaging fAPAR readings derived from the linear irradiance sensor orientated across and along the plant row provided an R² = 0.81 correlation with above-canopy NDVI. Across row sensor orientation also gave a similar correlation of R² = 0.76 allowing the user to reduce sampling time.     

Developing a new Crop Circle active canopy sensor-based precision nitrogen management strategy for winter wheat in North China Plain

Active canopy sensor (ACS)—based precision nitrogen (N) management (PNM) is a promising strategy to improve crop N use efficiency (NUE). The GreenSeeker (GS) sensor with two fixed bands has been applied to improve winter wheat (Triticum aestivum L.) N management in North China Plain (NCP). The Crop Circle (CC) ACS-470 active sensor is user configurable with three wavebands. The objective of this study was to develop a CC ACS-470 sensor-based PNM strategy for winter wheat in NCP and compare it with GS sensor-based N management strategy, soil N_min test-based in-season N management strategy and conventional farmer’s practice. Four site-years of field N rate experiments were conducted from 2009 to 2013 to identify optimum CC vegetation indices for estimating early season winter wheat plant N uptake (PNU) and grain yield in Quzhou Experiment Station of China Agricultural University located in Hebei province of NCP. Another nine on-farm experiments were conducted at three different villages in Quzhou County in 2012/2013 to evaluate the performance of the developed N management strategy. The results indicated that the CC ACS-470 sensor could significantly improve estimation of early season PNU (R² = 0.78) and grain yield (R² = 0.62) of winter wheat over GS sensor (R² = 0.60 and 0.33, respectively). All three in-season N management strategies achieved similar grain yield as compared with farmer’s practice. The three PNM strategies all significantly reduced N application rates and increased N partial factor productivity (PFP) by an average of 61–67 %. It is concluded that the CC sensor can improve estimation of early season winter wheat PNU and grain yield as compared to the GS sensor, but the PNM strategies based on these two sensors perform equally well for improving winter wheat NUE in NCP. More studies are needed to further develop and evaluate these active sensor-based PNM strategies under more diverse on-farm conditions.     

Mobile terrestrial laser scanner applications in precision fruticulture/horticulture and tools to extract information from canopy point clouds

LiDAR sensors are widely used in many areas and, in recent years, that includes agricultural tasks. In this work, a self-developed mobile terrestrial laser scanner based on a 2D light detection and ranging (LiDAR) sensor was used to scan an intensive olive orchard, and different algorithms were developed to estimate canopy volume. Canopy volume estimations derived from LiDAR sensor readings were compared to conventional estimations used in fruticulture/horticulture research and the results prove that they are equivalent with coefficients of correlation ranging from r = 0.56 to r = 0.82 depending on the algorithms used. Additionally, tools related to analysis of point cloud data from the LiDAR-based system are proposed to extract further geometrical and structural information from tree row crop canopies to be offered to farmers and technical advisors as digital raster maps. Having high spatial resolution information on canopy geometry (i.e., height, width and volume) and on canopy structure (i.e., light penetrability, leafiness and porosity) may result in better orchard management decisions. Easily obtainable, reliable information on canopy geometry and structure may favour the development of decision support systems either for irrigation, fertilization or canopy management, as well as for variable rate application of agricultural inputs in the framework of precision fruticulture/horticulture.     

Uniformity of wheat yield and quality using sensor assisted application of nitrogen

Proximal sensing, or obtaining information from close range, is a potentially useful tool for measuring the crop nitrogen status in real-time The objective of this study was to use proximal sensing of crop canopy spectral reflectance to evaluate variable-rate application of nitrogen in terms of its effect on yield and grain quality of winter wheat (Triticum aestivum L.). The sensor used was the Hydro-Precise N-Sensor System. Yield and grain quality maps were used as a basis for full-scale field trials with winter wheat growing under four nitrogen application treatments: a large (274 kg ha^?1), recommended (167 kg ha^?1) and two sensor-assisted (167 kg ha^?1) rates. The recommended rate of 167 kg N ha^?1 was given in a three-split application that meets the present Danish regulations to reduce nitrogen leaching. These require arable farmers to decrease nitrogen fertilizer application to 90% of the economically optimal level. Each farm’s baseline is calculated to take into account land quality, land allocated to each crop, and crop rotation. In the two sensor-assisted applications the Hydro-Precise N-Sensor System directs the last two of the three-split N application. Grain samples were collected directly from the grain flow of a combine harvester and analysed for protein, water and starch content. Grain data were related to and compared with combine yield meter registrations. Within the field, the variances of protein yield (698–1208 kg ha^?1) and grain protein (9.5–13.4%) were large. The nitrogen application treatments affected the average protein content (10.5–12.3%) and grain yield (9.87–10.42 t ha^?1) strongly. The grain starch content was largest in the uniform and sensor applied systems and smallest in the high nitrogen application treatment. Applying nitrogen according to the Hydro-Precise N-Sensor System did not increase grain yield or the protein and starch contents. Minor differences only were observed in both protein content and yield between uniform-rate N application and sensor-based variable-rate N application.     

Identification of yellow rust in wheat using in-situ spectral reflectance measurements and airborne hyperspectral imaging

The aim of this study was to evaluate the accuracy of the spectro-optical, photochemical reflectance index (PRI) for quantifying the disease index (DI) of yellow rust (Biotroph Puccinia striiformis) in wheat (Triticum aestivum L.), and its applicability in the detection of the disease using hyperspectral imagery. Over two successive seasons, canopy reflectance spectra and disease index (DI) were measured five times during the growth of wheat plants (3 varieties) infected with varying amounts of yellow rust. Airborne hyperspectral images of the field site were also acquired in the second season. The PRI exhibited a significant, negative, linear, relationship with DI in the first season (r ² = 0.91, n = 64), which was insensitive to both variety and stage of crop development from Zadoks stage 3–9. Application of the PRI regression equation to measured spectral data in the second season yielded a coefficient of determination of r ² = 0.97 (n = 80). Application of the same PRI regression equation to airborne hyperspectral imagery in the second season also yielded a coefficient of determination of DI of r ² = 0.91 (n = 120). The results show clearly the potential of PRI for quantifying yellow rust levels in winter wheat, and as the basis for developing a proximal, or airborne/spaceborne imaging sensor of yellow rust in fields of winter wheat.     

Mobile low-cost 3D camera maize crop height measurements under field conditions

To tackle global challenges such as food supply and renewable energy provision, the improvement of efficiency and productivity in agriculture is of high importance. Site-specific information about crop height plays an important role in reaching these goals. Crop height can be derived with a variety of approaches including the analysis of three-dimensional (3D) geodata. Crop height values derived from 3D geodata of maize (1.88 and 2.35 m average height) captured with a low-cost 3D camera were examined. Data were collected with a unique measurement setup including the mobile mounting of the 3D camera, and data acquisition under field conditions including wind and sunlight. Furthermore, the data were located in a global co-ordinate system with a straightforward approach, which can strongly reduce computational efforts and which can subsequently support near real-time data processing in the field. Based upon a comparison between crop height values derived from 3D geodata captured with the low-cost approach, and high-end terrestrial laser scanning reference data, minimum RMS and standard deviation values of 0.13 m (6.91% of average crop height), and maximum R² values of 0.79 were achieved. It can be concluded that the crop height measurements derived from data captured with the introduced setup can provide valuable input for tasks such as biomass estimation. Overall, the setup is considered to be a valuable extension for agricultural machines which will provide complementary crop height measurements for various agricultural applications.     

Simulating field-scale variability and precision management with a 3D hydrologic cropping systems model

Effective variable-rate nitrogen (N) management requires an understanding of temporal variability and field-scale spatial interactions (e.g. lateral redistribution of nutrients). Modeling studies, in conjunction with field data, can improve process understanding of agricultural management. CropSyst-Microbasin (CS-MB) is a fully distributed, 3-dimensional hydrologic cropping systems model that simulates small (10 s of hectares) heterogeneous agricultural watersheds with complex terrain. This study used a highly instrumented 10.9 ha watershed in the Inland Pacific Northwest, USA, to: (1) assess the accuracy of CS-MB simulations of field-scale variability in water transport and crop yield in comparison to observed field data, and (2) quantify differences in simulated yield and farm profitability between variable-rate and uniform fertilizer applications in low, average and high precipitation treatments. During water years 2012 and 2013 (a “water year” refers to October 1st through the following September 30th, where a given water year is named for the calendar year on September 30th), the model simulated surface runoff with a Nash–Sutcliffe efficiency (NSE) of 0.7, periodic soil water content (comparison to seasonal soil core measurements) with a root mean square error (RMSE) ≤0.05 m³ m^?3, and continuous soil water content (comparison to in situ soil sensors) at 15 of 20 microsites with NSE ≥0.4. The model predicted 2013 field variability in winter wheat yield with RMSE of 1100 kg ha^?1. Simulated uniform N management resulted in 0–35 kg ha^?1 greater field average yield in comparison to variable-rate management. The savings in fertilizer costs under variable-rate N management resulted in $23–$32 ha^?1 greater field average returns to risk. This study demonstrated the capacity of CS-MB to further understanding of simulated and observed field-scale spatial variability and simulated crop response to low, medium and high annual precipitation.     

Estimation of water productivity in winter wheat using the AquaCrop model with field hyperspectral data

Water productivity (WP) is a key element of agricultural water management in agricultural irrigated regions. The objectives of this study were: (i) to estimate biomass of winter wheat using spectral indices; (ii) integrate the estimation of biomass data with the AquaCrop model using a lookup table for higher accuracy biomass simulation; (iii) show estimation accuracy of the data assimilation method in yield and WP. Spectral variables and concurrent biomass, yield and WP of samples were acquired at the Xiaotangshan experimental site in Beijing, China, during the 2008/2009, 2009/2010, 2010/2011 and 2011/2012 winter wheat growing seasons. The results showed that all spectral indices had a highly significant relationship with biomass, especially normalized difference matter index, with R² and RMSE values of 0.84 and 1.43 t/ha, respectively. Simulation of biomass and yield by the AquaCrop model were in good agreement with the measured biomass and yield of winter wheat. The results showed that the data assimilation method (R² = 0.79 and RMSE = 0.12 kg/m³) could be used to estimate WP. The result indicated that the AquaCrop model could be used to estimate yield and WP with the aid of remote sensing for improving agricultural water resources management.     

NDVI variation according to the time of measurement,sampling size,positioning of sensor and water regime in different soybean cultivars

Although the information on the Normalized Difference Vegetation Index (NDVI) in plants under water deficit is often obtained from sensors attached to satellites, the increasing data acquisition with portable sensors has wide applicability in agricultural production because it is a fast, nondestructive method, and is less prone to interference problems. Thus, we carried out a set of experiments to investigate the influence of time, spatial plant arrangements, sampling size, height of the sensor and water regimes on NDVI readings in different soybean cultivars in greenhouse and field trials during the crop seasons 2011/12, 2012/13 and 2013/14. In experiments where plants were always evaluated under well-watered conditions, we observed that 9 a.m. was the most suitable time for NDVI readings regardless of the soybean cultivar, spatial arrangement or environment. Furthermore, there was no difference among NDVI readings in relation to the sampling size, regardless of the date or cultivar. We also observed that NDVI tended to decrease according to the higher height of the sensor in relation to the canopy top, with higher values tending to be at 0.8 m, but with no significant difference relative to 1.0 m—the height we adopted in our experiments. When different water regimes were induced under field conditions, NDVI readings measured at 9 a.m. by using a portable sensor were successful to differentiate soybean cultivars with contrasting responses to drought.     

Distinguishing nitrogen deficiency and fungal infection of winter wheat by laser-induced fluorescence

One of the most important tasks in precision farming is the site-specific application of fertilisers and pesticides in heterogeneous large-area fields. For such site-specific crop management, effective remote sensing methods for the detection of crop diseases and nutrient deficiencies are required. The aim of the present work was to compare laser-induced fluorescence (LIF) parameters from nitrogen-deficient and pathogen (rust and mildew)-infected winter wheat (Triticum aestivum L.) plants and to assess the potential of LIF to detect and discriminate between these types of stress. Both long term nitrogen deficiency and pathogen infection resulted in an increase of the ratio of fluorescence at 686 and 740 nm (F686/F740) accompanied by a reduction of leaf chlorophyll content to approximately 35 μg cm⁻². A linear negative correlation between chlorophyll content and F686/F740 ratio (r² = 0.78) was found for leaves with chlorophyll content ranging between 17 and 52 μg cm⁻². Since chlorophyll breakdown appeared an unspecific symptom to both nitrogen deficiency and pathogen infection, it was not possible to discriminate between these types of stress only by means of the F686/F740 ratio. Specific for the pathogen-infected leaves was a large heterogeneity in the records of their spectral parameters caused by inhomogeneous, discrete lesions of fungi infection. Nitrogen-deficient plants with homogeneous reduction in chlorophyll content showed, in contrast, more uniform readings of the spectral parameters. Thus, mildew- and rust-infected plants, grown under sufficient nitrogen fertilisation could be distinguished from those grown under reduced nitrogen supply by the higher variance of their spectral readings. The simultaneous scanning multipoint mode measurements of LIF and laser light reflection characteristics with parallel estimation of their heterogeneity is proposed for the discrimination between nitrogen deficiency and pathogen infection under field conditions.     

Estimation of bioenergy crop yield and N status by hyperspectral canopy reflectance and partial least square regression

The objective of this study was to compare performance of partial least square regression (PLSR) and best narrowband normalize nitrogen vegetation index (NNVI) linear regression models for predicting N concentration and best narrowband normalize different vegetation index (NDVI) for end of season biomass yield in bioenergy crop production systems. Canopy hyperspectral data was collected using an ASD FieldSpec FR spectroradiometer (350–2500 nm) at monthly intervals in 2012 and 2013. The cropping systems evaluated in the study were perennial grass {mixed grass [50 % switchgrass (Panicum virgatum L.), 25 % Indian grass “Cheyenne” (Sorghastrum nutans (L.) Nash) and 25 % big bluestem “Kaw” (Andropogon gerardii Vitman)] and switchgrass “Alamo”} and high biomass sorghum “Blade 5200” (Sorghum bicolor (L.) Moench) grown under variable N applications rates to estimate biomass yield and quality. The NNVI was computed with the wavebands pair of 400 and 510 nm for the high biomass sorghum and 1500 and 2260 nm for the perennial grass that were strongly correlated to N concentration for both years. Wavebands used in computing best narrowband NDVI were highly variable, but the wavebands from the red edge region (710–740 nm) provided the best correlation. Narrowband NDVI was weakly correlated with final biomass yield of perennial grass (r² = 0.30 and RMSE = 1.6 Mg ha^?1 in 2012 and r² = 0.37 and RMSE = 4.0 Mg ha^?1, but was strongly correlated for the high biomass sorghum in 2013 (r² = 0.72 and RMSE = 4.6 Mg ha^?1). Compared to the best narrowband VI, the RMSE of the PLSR model was 19–41 % lower for estimating N concentration and 4.2–100 % lower for final biomass. These results indicates that PLSR might be best for predicting the final biomass yield using spectral sample obtained in June to July, but narrowband NNVI was more robust and useful in predicting N concentration.     

Optimization of corn plant population according to management zones in Southern Brazil

Precision agriculture relies on site-specific interventions determined by the spatial variability of factors driving plant growth. The main objective of this study was to assess the efficiency of variable-rate seeding of corn (Zea mays L.) with delineated management zones. This study involved two experiments carried out in Não-Me-Toque, Rio Grande do Sul, Brazil. For the first experiment, carried out in 2009/2010, management zones were delineated by the farmer’s knowledge of the crop field. The field was split into low (LZ), medium (MZ) and high (HZ) crop performance zones. In the second experiment, carried out in 2010/2011, management zones were delineated by overlaying standardized yield data from nine crop seasons (seven of soybean and two of corn). The experiment was carried out with a randomized block design with three management zones and five corn seeding rates ranging from 50 000 to 90 000 seeds per ha^?1. The soil was a Rhodic Hapludox with a subtropical climate. Optimization of the corn plant population within the field increased grain yield compared to the reference plant population (70 000 plants ha^?1). Yield increases in the LZ, due to corn plant population reduction in relation to the target population, were 1.20 and 1.90 Mg ha^?1 for first and second experiments, respectively. This resulted in economic gains of 19.8 and 28.7 %, respectively. Yield increases in the HZ were 0.89 and 0.94 Mg ha^?1, respectively, and were due to an increase in plant population in relation to the target population. This resulted in economic gains of 5.6 and 6.6 % for the first and second experiments, respectively. In the MZ, the adjustment of the target plant population was not necessary. Optimizing corn population according to management zones is a promising tool for precision agriculture in Southern Brazil.     

Determining the within-field yield variability from seasonally changing soil conditions

Crop yield variations are strongly influenced by the spatial and temporal availabilities of water and nitrogen in the soil during the crop growth season. To estimate the quantities and distributions of water and nitrogen within a given soil, process-oriented soil models have often been used. These models require detailed information about the soil characteristics and profile architecture (e.g., soil depth, clay content, bulk density, field capacity and wilting point), but high resolution information about these soil properties, both vertically and laterally, is difficult to obtain through conventional approaches. However, on-the-go electrical resistivity tomography (ERT) measurements of the soil and data inversion tools have recently improved the lateral resolutions of the vertically distributed measurable information. Using these techniques, nearly 19,000 virtual soil profiles with defined layer depths were successfully created for a 30 ha silty cropped soil over loamy and sandy substrates in Central Germany, which were used to initialise the CArbon and Nitrogen DYnamics (CANDY) model. The soil clay content was derived from the electrical resistivity (ER) and the collected soil samples using a simple linear regression approach (the mean R² of clay = 0.39). The additional required structural and hydrological properties were derived from pedotransfer functions. The modelling results, derived soil texture distributions and original ER data were compared with the spatial winter wheat yield distribution in a relatively dry year using regression and boundary line analysis. The yield variation was best explained by the simulated soil water content (R² = 0.18) during the grain filling and was additionally validated by the measured soil water content with a root mean square error (RMSE) of 7.5 Vol%.     

Evaluation of spectral indices and continuous wavelet analysis to quantify aphid infestation in wheat

Wheat aphid, Sitobion avenae F. is one of the most destructive insects infesting winter wheat and appears almost annually in northwest China. Past studies have demonstrated the potential of remote sensing for detecting crop diseases and insect damage. This study aimed to investigate the spectroscopic estimation of leaf aphid density by applying continuous wavelet analysis to the reflectance spectra (350–2 500 nm) of 60 winter wheat leaf samples. Continuous wavelet transform (CWT) was performed on each of the reflectance spectra to generate a wavelet power scalogram compiled as a function of wavelength location and scale of decomposition. Linear regression between the wavelet power and aphid density was to identify wavelet features (coefficients) that might be the most sensitive to aphid density. The results identified five wavelet features between 350 and 2 500 nm that provided strong correlations with leaf aphid density. Spectral indices commonly used to monitor crop stresses were also employed to estimate aphid density. Multivariate linear regression models based on six sensitivity spectral indices or five wavelet features were established to estimate aphid density. The results showed that the model with five wavelet features (R² = 0.72, RMSE = 16.87) performed better than the model with six sensitivity spectral indices (R² = 0.56, RMSE = 21.19), suggesting that the spectral features extracted through CWT might potentially reflect aphid density. The results also provided a new method for estimating aphid density using remote sensing.     

Autonomous field navigation,data acquisition and node location in wireless sensor networks

To overcome the limited transmission range of spatially separated nodes of a wireless sensor network (WSN), a small 4-wheel autonomous robot assembled the data from nodes distributed in a vineyard. First, the robot followed a predefined way-point route between the grapevine rows, in order to evaluate the sensor node locations by their received signal strength indication (RSSI). Then, the recorded and geo-referenced RSSI data were analysed and mapped. By using the evaluated node positions, an optimised second route was generated. While navigating, a laser scanner was used for obstacle detection and avoidance. Path planning with known positions of the nodes reduced the driving time by 15 times compared with the first run, because the hybrid control system used was capable of navigating within the plantation even perpendicular to the row structures. For locating the nodes, results based on trilateration were compared with the values of an attached differential global navigation satellite system (DGNSS). The results showed that it is possible to locate and geo-reference the sensor nodes with a robot, even without any prior knowledge about their absolute position. The best achieved location showed a deviation with DGNSS of 1.2 m and with RSSI trilateration of 0.6 m compared to the actual position.     

Development of an in-season estimate of yield potential utilizing optical crop sensors and soil moisture data for winter wheat

When utilizing optical sensors to make in-season agronomic recommendations in winter wheat, one parameter often required is the in-season grain yield potential at the time of sensing. Current estimates use an estimate of biomass, such as normalized difference vegetation index (NDVI), and growing degree days (GDDs) from planting to NDVI data collection. The objective of this study was to incorporate soil moisture data to improve the ability to predict final grain yield in-season. Crop NDVI, GDDs that were adjusted based upon if there was adequate water for crop growth, and the amount of soil profile (0–0.80 m) water were incorporated into a multiple linear regression model to predict final grain yield. Twenty-two site-years of N fertility trials with in-season grain yield predictions for growth stages ranging from Feekes 3 to 10 were utilized to calibrate the model. Three models were developed: one for all soil types, one for loamy soil textured sites, and one for coarse soil textured sites. The models were validated with 11 independent site-years of NDVI and weather data. The results indicated there was no added benefit to having separate models based upon soil types. Typically, the models that included soil moisture, more accurately predicted final grain yield. Across all site years and growth stages, yield prediction estimates that included soil moisture had an R² = 0.49, while the current model without a soil moisture adjustment had an R² = 0.40.