Frédéric Cointault

Measurement of the Motion of Fertilizer Particles Leaving a Centrifugal Spreader Using a Fast Imaging System

Although mechanically simple, centrifugal spreaders used for mineral fertilization involve complex physics that cannot be fully characterized at the present time. We are developing sensors to evaluate the spatial distribution of the fertilizer on the ground based on the measurement of initial flight conditions of fertilizer granules after their ejection by the spreading disk. The techniques developed are based on the analysis of images of the area around the disk showing the granule ejection. A high resolution – low cost imaging system for the analysis of high speed particle projection developed for this specific purpose is presented in this paper. The system, based on a camera and a sequence of flashes, is used to characterize the centrifugal spreading of fertilizer particles ejected at speeds of approximately 30m s−1. It automatically computes the direction of ejection and velocity of each granule observed in the image. Multi-exposure images collected with the camera installed perpendicular to the output flow of granules are analyzed to estimate the trajectories of the fertilizer granules, using different motion estimation methods.

3D Image Acquisition System Based on Shape from Focus Technique

This paper describes the design of a 3D image acquisition system dedicated to natural complex scenes composed of randomly distributed objects with spatial discontinuities. In agronomic sciences, the 3D acquisition of natural scene is difficult due to the complex nature of the scenes. Our system is based on the Shape from Focus technique initially used in the microscopic domain. We propose to adapt this technique to the macroscopic domain and we detail the system as well as the image processing used to perform such technique. The Shape from Focus technique is a monocular and passive 3D acquisition method that resolves the occlusion problem affecting the multi-cameras systems. Indeed, this problem occurs frequently in natural complex scenes like agronomic scenes. The depth information is obtained by acting on optical parameters and mainly the depth of field. A focus measure is applied on a 2D image stack previously acquired by the system. When this focus measure is performed, we can create the depth map of the scene.

The Use of High-Speed Imaging Systems for Applications in Precision Agriculture

High speed imaging (HSI) has been widely used for industrial and military applications such as ballistics, hypervelocity impact, car crash studies, fluid mechanics, and others. In agriculture HSI is mainly used in two domains that both require fast processing: fertilization and spraying.  Fertilization, be it organic or mineral, is essential to agriculture. Over-fertilization can reduce yield and lead to environmental pollution (Mulligan et al., 2006). To prevent these consequences, the fertilization process must be controlled. In Europe and worldwide, mineral fertilization is performed using centrifugal spreaders because they are more cost-efficient than pneumatic spreaders. The process of centrifugal spreading is based on spinning discs which eject large numbers of grains at high speeds (30 to 40 ms-1). To control the spreading process and to predict the distribution pattern on the soil, several characteristics need to be accurately evaluated, i.e., ejection parameters such as velocity and direction, plus granulometry and the angular distribution.  The spray quality generated by agricultural nozzles plays an important role in the application of plant protection products. The ideal nozzle-pressure combination should maximize spray efficiency by increasing deposition and transfer of a lethal dose to the target (Smith et al., 2000) while minimizing residues (Derksen et al., 2008) and off-target losses such as spray drift (Nuyttens et al., 2007a) and user exposure (Nuyttens et al., 2009a). The most important spray characteristics influencing the efficiency of the pesticide application process are the droplet sizes, the droplet velocities and directions, the volume distribution pattern, the spray sheet structure and length, the structure of

Texture Classification with Generalized Fourier Descriptors in Dimensionality Reduction Context: An Overview Exploration

In the context of texture classification, this article explores the capacity and the performance of some combinations of feature extraction, linear and nonlinear dimensionality reduction techniques and several kinds of classification methods. The performances are evaluated and compared in term of classification error. In order to test our texture classification protocol, the experiment carried out images from two different sources, the well known Brodatz database and our leaf texture images database.

Plant leaf roughness analysis by texture classification with generalized Fourier descriptors in a dimensionality reduction context

In the context of plant leaf roughness analysis for precision spraying, this study explores the capability and the performance of some combinations of pattern recognition and computer vision techniques to extract the roughness feature. The techniques merge feature extraction, linear and nonlinear dimensionality reduction techniques, and several kinds of methods of classification. The performance of the methods is evaluated and compared in terms of the error of classification. The results for the characterization of leaf roughness by generalized Fourier descriptors for feature extraction, kernel-based methods such as support vector machines for classification and kernel discriminant analysis for dimensionality reduction were encouraging. These results pave the way to a better understanding of the adhesion mechanisms of droplets on leaves that will help to reduce and improve the application of phytosanitary products and lead to possible modifications of sprayer configurations.

Texture or Color Analysis in Agronomic Images for Wheat Ear Counting

In agronomy, image processing techniques are more and more used to detect crop, weeds, diseases ... We proposed to study the feasibility to use color and/or texture analysis to evaluate the number of wheat ears per m2 to simplify the manual countings currently done. In this paper we present firstly the use of color and texture image processing together to detect the ears, before to propose and compare different texture image segmentation techniques based on feature extraction by first and higher order statistical methods. The extracted features are used for unsupervised pixel classification to obtain the different classes in the image, before to use the k-means algorithm. Three methods have been tested with very heterogeneous results, except the run length technique for which the results are close to the manual countings (66% error). The hypothesis took into account for the textural analysis methods are currently modify to justify them more accurately, especially concerning the number of classes and the size of the analysis window.

Spray Droplet Characterization from a Single Nozzle by High Speed Image Analysis Using an In-Focus Droplet Criterion

Accurate spray characterization helps to better understand the pesticide spray application process. The goal of this research was to present the proof of principle of a droplet size and velocity measuring technique for different types of hydraulic spray nozzles using a high speed backlight image acquisition and analysis system. As only part of the drops of an agricultural spray can be in focus at any given moment, an in-focus criterion based on the gray level gradient was proposed to decide whether a given droplet is in focus or not. In a first experiment, differently sized droplets were generated with a piezoelectric generator and studied to establish the relationship between size and in-focus characteristics. In a second experiment, it was demonstrated that droplet sizes and velocities from a real sprayer could be measured reliably in a non-intrusive way using the newly developed image acquisition set-up and image processing. Measured droplet sizes ranged from 24 μm to 543 μm, depending on the nozzle type and size. Droplet velocities ranged from around 0.5 m/s to 12 m/s. The droplet size and velocity results were compared and related well with the results obtained with a Phase Doppler Particle Analyzer (PDPA).

Using Image Texture and Spectral Reflectance Analysis to Detect Yellowness and Esca in Grapevines at Leaf-Level

Plant diseases are one of the main reasons behind major economic and production losses in the agricultural field. Current research activities enable large fields monitoring and plant disease detection using innovative and robust technologies. French grapevines have a reputation for producing premium quality wines, however, these major fruit crops are susceptible to many diseases, including Esca, Downy mildew, Powdery mildew, Yellowing, and many others. In this study, we focused on two main infections (Esca and Yellowing), and data were gathered from fields that were located in Aquitaine and Burgundy regions, France. Since plant diseases can be diagnosed from the properties of the leaf, we acquired both Red-Green-Blue (RGB) digital image and hyperspectral reflectance data from infected and healthy leaves. Biophysical parameters that were produced by the PROSPECT model inversion together with texture parameters compiled from the literature were deduced. Then we investigated their relationship to damage caused by Yellowing and Esca. This study examined whether spectral and textural data can identify the two diseases through the use of Neural Networks. We obtained an overall accuracy of 99% for both of the diseases when textural and spectral data are combined. These results suggest that, first, biophysical parameters present a valid dimension reduction tool that could replace the use of complete hyperspectral data. Second, remote sensing using spectral reflectance and digital images can make an overall nondestructive, rapid, cost-effective, and reproducible technique to determine diseases in grapevines with a good level of accuracy.

Development of Spectral Disease Indices for ‘Flavescence Dorée’ Grapevine Disease Identification

Spectral measurements are employed in many precision agriculture applications, due to their ability to monitor the vegetation’s health state. Spectral vegetation indices are one of the main techniques currently used in remote sensing activities, since they are related to biophysical and biochemical crop variables. Moreover, they have been evaluated in some studies as potentially beneficial for detecting or differentiating crop diseases. Flavescence Dorée (FD) is an infectious, incurable disease of the grapevine that can produce severe yield losses and, hence, compromise the stability of the vineyards. The aim of this study was to develop specific spectral disease indices (SDIs) for the detection of FD disease in grapevines. Spectral signatures of healthy and diseased grapevine leaves were measured with a non-imaging spectro-radiometer at two infection severity levels. The most discriminating wavelengths were selected by a genetic algorithm (GA) feature selection tool, the Spectral Disease Indices (SDIs) are designed by exhaustively testing all possible combinations of wavelengths chosen. The best weighted combination of a single wavelength and a normalized difference is chosen to create the index. The SDIs are tested for their ability to differentiate healthy from diseased vine leaves and they are compared to some common set of Spectral Vegetation Indices (SVIs). It was demonstrated that using vegetation indices was, in general, better than using complete spectral data and that SDIs specifically designed for FD performed better than traditional SVIs in most of cases. The precision of the classification is higher than 90%. This study demonstrates that SDIs have the potential to improve disease detection, identification and monitoring in precision agriculture applications.

ROUGHNESS EVALUATION OF VINE LEAF BY IMAGE PROCESSING

The study of leaf surface roughness is very important in the domain of precision spraying. It is one of the parameters that allow to reduce costs and losses of phytosanitary prod- ucts and to improve the spray accuracy. Moreover, the leaf roughness is related to adhesion mechanisms of liquid on a surface. It can be used to define leaf nature surface (hy- drophilic/hydrophobic). The main goal of this study is thus to estimate and to follow the evolution of leaf roughness using image processing and computer vision. The develop- ment and application of computer vision for measurement of surface leaf roughness using artificial neural networks will be described. The system for image acquisition of leaf surface consists of scanning electron microscope (SEM). The images of leaf surface are captured and analyzed to estimate the optical roughness. 2-D Fast Fourier Trans- form (FFT) algorithm and Co-occurrence Matrix are used for texture analysis. A multilayer perceptron (MLP) neural network is used to model and predict the optical roughness values.