Disease detection and control is thus one of the main objectives of vineyard research in France. Monitoring diseases manually is fastidious and time consuming, so current research aims to develop an automatic detection of vineyard diseases. This project explored the use of a high-resolution multi-spectral camera embedded on a UAV (Unmanned Aerial Vehicle) to identify the infected zones in a field. In-field spectrometry studies were performed to identify the best spectral bands for the sensor design. The best models were found to be the function of the grapevine variety considered and the 520-600-650-690-730-750-800 nm bands were found to be the most efficient for all types of grapevines, with an overall classification accuracy of more than 94%.

In strawberry production, a balanced and accurate irrigation schedule is essential, because of the high sensitivity of strawberry plants to water deficits and waterlogging. The optimal irrigation management strategy can, however, only be obtained by an accurate crop monitoring system. To replace the current visual inspection methods, which are subjective, time consuming and labour-intensive, the performance of the COmpact hyperSpectral Imaging system (COSI) mounted on an RPAS (Remotely Piloted Aircraft System) was evaluated. The study, focusing on different irrigation treatments in strawberry cultivation, unraveled the potential of the COSI system, to monitor field variations, even at small scale. Growth inhibition and differences in plant physiology due to water deficit could be detected. As such, the COSI system has shown potential for steering irrigation management decisions in strawberry cultivation.

The objective of this work was to determine the most appropriate growth stage to make reflectance measurements that would indicate yield in high yielding winter barley crops. The results indicated that where different rates of fertiliser N were applied, at the same crop growth stages, the best relationship between vegetation indices, calculated on the basis of reflectance measurements, and grain yield were found to occur from booting to early grain fill. Where the timing of fertiliser N inputs was different, for a given level of fertiliser N addition, poor correlations between vegetation indices and grain yield during the stem elongation phase were observed.

This paper focuses on maturity evaluation derived by a color camera for a sweet pepper robotic harvester. Different color and morphological features for sweet pepper maturity were evaluated. Side view and bottom view of sweet paper were analyzed and compared for their ability to classify into 4 maturity classes. The goal of this study was to differentiate between the two center classes which are difficult to separate. Statistical analysis of 13 different features in reliance to the maturity classification and the views indicated the best features for classification. The results show that the features that can be used for classification between the two central classes from both bottom and side views are: Hue range, Equal2Real – the ratio between the equivalent equal sized circle perimeter to the shape perimeter and Area2Peri – the ratio between the area to the perimeter.

Potassium (K) is a macronutrient which plays a vital role on crop growth and metabolism. After N the requirements for K are greatest for most arable crops and so the availability of K is of critical importance to optimise production. The precision nutrient management of arable crops requires accurate and timely assessment of crop nutrient status. Much research and practice has focused on crop N status, while there has been a lack of focus on other important nutrients such as K. Therefore, in this study we assess the robustness of 12 fluorescence channels and several indices to predict nutrient status (K, Mg and Ca) across two cereal crops with different row management and lime status on an acidic K deficient soil. A multi-factorial experiment was used with the following treatment factors: crop (barley, wheat), K fertilizer rates (0, 25, 50, 100 kg K/ ha), lime (nil, 1 t/ ha) and two management factors (inter-row, windrow). At flowering the crop was sampled for biomass and nutrient content and proximal sensing (using a Multiplex fluorometer) undertaken of the crop canopy. Crop variables showed significant treatment effects. For instance, all crop variables were greater under the windrow treatment than the inter-row, K rate significantly increased grain yield and TGW, but K rate decreased protein and grain Ca and Mg content, also the grain yield was significantly greater under lime compared with the nil treatment. These crop effects enabled the identification of significant crop-fluorescence relationships. For instance, SFR_R (a chlorophyll index) predicted crop biomass (regardless of crop species) and FLAV predicted with the grain protein of windrow-grown barley. These results are promising and suggest crop-fluorescence relationships can be used to inform crop nutrient status which could be used to aid management decisions. Thus, there is good potential for fluorescence sensing to quantify crop K status and the opportunity to improve the timing and precision of K management for application within a precision agriculture system.

LiDAR technology is one option to collect spatial data about canopy geometry in many crops. However, the method of data acquisition includes many errors related to the LiDAR sensor, the GNSS receiver and the data acquisition set up. Therefore, the objective of this study was to evaluate the errors involved in the data acquisition from a mobile terrestrial laser scanner (MTLS). Regular shaped objects were scanned with a developed MTLS in two different tests: i) with the system mounted on a vehicle and ii) with the system mounted on a platform running over a rail. The errors of area estimation varied between 0.001 and 0.071 m2 for the circle, square and triangle objects. The errors on volume estimations were between 0.0003 and 0.0017 m3, for cylinders and truncated cone.

An accurate 3D model of an outdoor scene can be used in many different scenarios of precision agriculture, for instance to analyse the silhouette of a tree crown canopy for precision spraying, to count fruit for fruit yield prediction or to simply navigate a vehicle between the plant rows. Instead of using stereovision, limited by the problems of different light intensities, or by using expensive multi-channel 3D range finder (LIDAR scanner), limited by the number of channels, this work investigates the possibility of using two single channel LIDAR scanners mounted on a small robot to allow a real-time 3D object reconstruction of the robot environment. The approach used readings captured by two LIDAR scanners, SICK LMS111 and SICK TiM310, where the first one was scanning horizontally and the second one vertically. In order to correctly map the 3D points of the readings from the vertical sensor into a 3D space, a custom SLAM algorithm based on image registration techniques was used to calculate the new positions of the robot. The approach was tested in an indoor and outdoor environment, proving its accuracy with an error rate of 0.02 m±0.02 m for vertical and −0.01 m±0.13 m for the horizontal plane.

An increasing number of farm machines nowadays implement precision agriculture technologies. Most of these operate through proximal sensing using optical sensors (i.e. NIR or Vis-NIR). Imaging techniques in this context have received minor consideration due to the complex analysis of the data but on the other side offer great flexibility. This study reports a preliminary pilot imaging multi-sensor retrofit system to be applied independently on a wide range of agricultural machines and able to test different monitoring or control image-based applications for precision agriculture. The process, based on RGB image, was tested for in-field discrimination of weeds in lettuce and broccoli crops. It works by discriminating and extracting single plants from the soil and weeds. However, to be truly implementable, the experimental code should be optimized in order to shorten the time needed for acquisition and processing.

In precision agriculture, the selection and use of appropriate sensors determine the type and quality of information that will feed decision-support models. A wide variety of sensors, spectral ranges, data collection and processing approaches are used, sometimes leading to confusion. Whether in transmission or reflectance mode, multispectral or hyperspectral, laboratory or field-based or even satellite-borne, in order to achieve meaningful and accurate measurements it is essential to have a clear understanding of which part of the electromagnetic spectrum the sensors relate to and how the corresponding radiation interacts with the substrate (e.g. soils, crops, livestock products). Sensors in the visible range (390-700 nm) use colour to identify certain properties of the substrate (e.g. chlorophyll and pigments in crops, organic matter contents in soil) and can be used to detect and quantify colour changes that could, in turn, be correlated with changes in those properties. Alternatively, radiation in the near (NIR, 750-2500 nm) and mid infrared (MIR, 2500-25 000 nm) interacts with the molecular bonds that constitute organic and inorganic matter and, therefore, sensors with detectors in these ranges provide different but interrelated information on the chemical composition of the substrate. Shorter wavelength radiation in the form of X-Rays (0.1-10 nm) induces fluorescence in the substrate and XRF sensors provide elemental atomic information that is highly applicable to the study of soils, sediments and fluids. At the James Hutton Institute, we have expertise in the use of all these types of sensors and are developing practical applications based on a thorough understanding of the processes involved. In this paper we provide an overview of the capabilities and applications of the different sensors used in precision agriculture, not only with a theoretical understanding, but also with an awareness of the practicalities involved.

Peanut (Arachis Hypogea) is one of the few major agronomic crops for which a yield monitor is not commercially available. This paper describes an ongoing project whose long-term goal is to adapt an optical sensor originally developed for cotton yield monitoring for use as a peanut yield monitor (PYM). The immediate objective of the work reported here was to evaluate the PYM under harvest conditions typical in southern Georgia, USA. The PYM consists of two mass-flow sensors, a data acquisition system, and a DGPS receiver. The PYM was evaluated on three fields totaling 29 ha during the 2016 harvest season. Percent error between the scale load and calculated load was 2% or better for the first field tested, but increased greatly for subsequent fields that were tested, most likely caused by damage to the sensor lens from the impact of pebbles.

The goal of this work was to cluster maize plants perception points under six different growth stages in noisy 3D point clouds with known positions. The 3D point clouds were assembled with a 2D laser scanner mounted at the front of a mobile robot, fusing the data with the precise robot position, gained by a total station and an Inertial Measurement Unit. For clustering the single plants in the resulting point cloud, a graph-cut based algorithm was used. The algorithm results were compared with the corresponding measured values of plant height and stem position. An accuracy for the estimated height of 1.55 cm and the stem position of 2.05 cm was achieved.

This paper presents an inspection system integrated into an on-ground autonomous platform with range of approximately 80 km. The vehicle is prepared to autonomously cover a field following a predefined route plan. Two types of cameras were integrated in the platform. An RGB-D sensor and a reflex camera were placed in a fixture and connected to a high-performance computer. The heterogeneous information acquired from the RGB-D was later integrated to automatically generate 3D maps of the crops by using custom software developed in the authors’ previous work. The inspection system performance was tested in actual vineyards by conducting several samplings in 2016. Results show that the proposed technology is viable and can provide complementary information to other inspection alternatives.

Due to the lack of reliable yield monitor for sugarcane, production factors which impact and limit stalk yield within fields are not well-known. Thus, this study aims to evaluate whether canopy sensor technology is able to identify sugarcane biomass variability and whether obtaining other agronomic variable data can assist on biomass quantification. For that, forty targeted plots were allocated within two sugarcane-producing fields and data consisted on manual biometric evaluation, aboveground biomass measurement and canopy reflectance. As an ongoing experiment, only two evaluations were addressed (~0.3 and 0.5 m stalk height). On the earliest stage, canopy sensor readings were correlated to sugarcane biomass and their sensitivity to biomass variability was high. Further, data collected on the first evaluation was efficient in predicting biomass amount after 30 days. On the second, canopy sensor readings effectiveness to predict biomass was reduced. These findings suggest that crop canopy reflectance sensing is a useful approach to investigate sugarcane biomass spatial-variability within fields on early stages.

Crop phenotype is usually expressed in terms of characteristics like plant height, leaf architecture and leaf area index (LAI). In the case of maize, stalk diameter is seldom quantified because its measurement does not readily lend itself to automation. Justification for automating the measurement of stalk diameter and plant spacing is based on the finding that stalk diameter was able to account for about 65% of the variability in maize yield per plant in three irrigated field studies. A high-speed reflectance sensor and simulation apparatus was developed to explore the potential for automating maize stalk diameter assessment. The prototyped system accurately measured both stalk diameter and plant spacing in the laboratory at simulated velocities up to 12 km/h.

The human population is expected to reach 9 billion by 2050 and thus high yield crop varieties need to be developed. Remote sensing can estimate crop parameters non-destructively and quickly. The aim of this study was to compare and evaluate the use of a commercial RGB camera with an expensive canopy sensor in the crop development of two legumes. The RGB camera based vegetation index (NGRDI) was compared with the canopy sensor derived vegetation indices (NDVI and NDRE) for estimating legume crop growth parameters. The results indicated that the use of a simple digital camera RGB can in some cases replace spectral canopy sensors.

The research concerns the changes of spectral characteristics of reflected radiation (360 to 1 000 nm) of spring wheat leaves under nitrogen deficiency and moderate soil drought. The efficiency of factorial influence (η2) on chlorophyll index was equal to 20% and 4% under nitrogen and water deficiency, respectively. Most significantly soil drought influenced the water index WRI (η2=55%) and the light diffusion index R800 (η2=28%), which was caused by changes in leaf structure. At low levels of nitrogen supply, these parameters did not change or changed only slightly (η2=2%). It may be deduced that the data base for crop monitoring in precision farming systems must contain a series of optical criteria for assessing specific and non-specific changes in optical characteristics of a crop canopy under the impact of various stress factors.

The use of plant growth regulators (PGRs) on winter wheat in the UK is a common practice. Significant yield benefits can be gained from its use. For the most efficient use of PGRs the rates of application must be matched to the crop canopy. The variation of crop growth was monitored during the spring growing season using vegetation indices derived from satellite imagery. Application rates of PGRs were matched to changes in biomass as measured by the imagery. Final yields were measured using a GPS combine yield meter. The variable application showed an average yield benefit of 0.4 t/ha over the uniform full recommended application rate.

The detection and identification of plant diseases is a fundamental task in sustainable crop production. An accurate estimate of disease incidence, disease severity and negative effects on yield quality and quantity is important for precision crop production, horticulture, plant breeding or fungicide screening as well as in basic and applied plant research. Particularly hyperspectral imaging of diseased plants offers insight into processes during pathogenesis. By hyperspectral imaging and subsequent data analysis routines, it was possible to realize an early detection, identification and quantification of different relevant plant diseases. Depending on the measuring scale, even subtle processes of defence and resistance mechanism of plants could be evaluated. Within this scope, recent results from studies in barley, wheat and sugar beet and their relevant foliar diseases will be presented.