Sylvain Jay

In-field crop row phenotyping from 3D modeling performed using Structure from Motion

We propose a phenotyping method to characterize the crop row structure.A crop row 3D model is built using a single camera and Structure from Motion.Structural parameters such as height and leaf area are estimated from this 3D model.Our method was tested with various plant structures and outdoor conditions.Strong agreements were obtained between estimated and actual heights and leaf areas. This article presents a method for crop row structure characterization that is adapted to phenotyping-related issues. In the proposed method, a crop row 3D model is built and serves as a basis for retrieving plant structural parameters. This model is computed using Structure from Motion with RGB images acquired by translating a single camera along the row. Then, to estimate plant height and leaf area, plant and background are discriminated by a robust method that uses both color and height information in order to handle low-contrasted regions. The 3D model is scaled and the plant surface is finally approximated using a triangular mesh.The efficacy of our method was assessed with two data sets collected under outdoor conditions. We also evaluated its robustness against various plant structures, sensors, acquisition techniques and lighting conditions. The crop row 3D models were accurate and led to satisfactory height estimation results, since both the average error and reference measurement error were similar. Strong correlations and low errors were also obtained for leaf area estimation. Thanks to its ease of use, estimation accuracy and robustness under outdoor conditions, our method provides an operational tool for phenotyping applications.

Estimationofwater column parameters with a maximum likelihood approach

In this article, we use a well-known reflectance model of a water column for estimating the model parameters (depth and concentrations of different water constituents) with a maximum likelihood approach. Tested on simulated data, the method performs well, especially for depths between a few meters and about 10m, and a SNR greater than 10dB. Moreover, we calculate the Cramér-Rao lower bounds in order to assess the performances of this estimation process. We show that the variances of the estimators come closer to these CRBs when the number of training pixels grows. Moreover, it turns out that the ML estimates of Cϕ, Ccdom and CNap are efficient even for low sample sizes.

Underwater target detection with hyperspectral remote-sensing imagery

This paper presents a new way of detecting underwater targets with hyperspectral remote-sensing data. The idea is to use a bathymetric model of subsurface reflectance to correct the spectral distortions due to water crossing. Then we derive the Matched filter (MF) from the Likelihood Ratio Test (LRT) built to decide whether the target is present or absent. Tested on both simulated and real images, this new detector appears to overcome classical filters in case of underwater targets. If the depth is unknown, it can be estimated using the maximum likelihood approach, and we show on simulations that detection performances are not very sensitive to the depth estimation accuracy.

A Spectral–Spatial Approach for Hyperspectral Image Classification Using Spatial Regularization on Supervised Score Image

This paper proposes a novel approach to classify hyperspectral (HS) images using both spectral and spatial information. It first consists of a supervised spectral dimension reduction step that transforms the HS image into a score image that has fewer channels. These channels are chosen so as to enhance distances between classes to be discriminated and to reduce background variability, thus leading to edges that correspond to actual class borders. In the second step, applying an edge-preserving spatial regularization on this score image leads to a lowered background variability. Therefore, in the third step, the pixel-wise classification of the regularized score image is greatly improved. We implement this approach using the partial least squares (PLS) method for spectral dimension reduction and the anisotropic diffusion for spatial regularization. We then compare linear discriminant analysis (LDA), K-nearest neighbors (KNN), and support vector machine (SVM) for the class decision. The effectiveness of our method was evaluated with three remotely sensed HS images. Its robustness was also assessed for different training sets, since the latter has a crucial influence on classification performance. On average, our method gave better results in terms of classification accuracy and was more robust than other classification methods tested with the same images.

Exploring the potential of PROCOSINE and close-range hyperspectral imaging to study the effects of fungal diseases on leaf physiology

The detection of plant diseases, including fungi, is a major challenge for reducing yield gaps of crops across the world. We explored the potential of the PROCOSINE radiative transfer model to assess the effect of the fungus Pseudocercospora fijiensis on leaf tissues using laboratory-acquired submillimetre-scale hyperspectral images in the visible and near-infrared spectral range. The objectives were (i) to assess the dynamics of leaf biochemical and biophysical parameters estimated using PROCOSINE inversion as a function of the disease stages, and (ii) to discriminate the disease stages by using a Linear Discriminant Analysis model built from the inversion results. The inversion results show that most of the parameter dynamics are consistent with expectations: for example, the chlorophyll content progressively decreased as the disease spreads, and the brown pigments content increased. An overall accuracy of 78.7% was obtained for the discrimination of the six disease stages, with errors mainly occurring between asymptomatic samples and first visible disease stages. PROCOSINE inversion provides relevant ecophysiological information to better understand how P. fijiensis affects the leaf at each disease stage. More particularly, the results suggest that monitoring anthocyanins may be critical for the early detection of this disease.

Predicting minimum uncertainties in the inversion of ocean color geophysical parameters based on Cramer-Rao bounds

We present an analytical approach based on Cramer-Rao Bounds (CRBs) to investigate the uncertainties in estimated ocean color parameters resulting from the propagation of uncertainties in the bio-optical reflectance modeling through the inversion process. Based on given bio-optical and noise probabilistic models, CRBs can be computed efficiently for any set of ocean color parameters and any sensor configuration, directly providing the minimum estimation variance that can be possibly attained by any unbiased estimator of any targeted parameter. Here, CRBs are explicitly developed using (1) two water reflectance models corresponding to deep and shallow waters, resp., and (2) four probabilistic models describing the environmental noises observed within four Sentinel-2 MSI, HICO, Sentinel-3 OLCI and MODIS images, resp. For both deep and shallow waters, CRBs are shown to be consistent with the experimental estimation variances obtained using two published remote-sensing methods, while not requiring one to perform any inversion. CRBs are also used to investigate to what extent perfect a priori knowledge on one or several geophysical parameters can improve the estimation of remaining unknown parameters. For example, using pre-existing knowledge of bathymetry (e.g., derived from LiDAR) within the inversion is shown to greatly improve the retrieval of bottom cover for shallow waters. Finally, CRBs are shown to provide valuable information on the best estimation performances that may be achieved with the MSI, HICO, OLCI and MODIS configurations for a variety of oceanic, coastal and inland waters. CRBs are thus demonstrated to be an informative and efficient tool to characterize minimum uncertainties in inverted ocean color geophysical parameters.

Near infrared spectra and hyperspectral imaging of medieval fortress walls in Carcassonne: a comprehensive interdisciplinary field study

A comprehensive study has been launched in the medieval fortress of Carcassonne involving a cooperation between the universities of Umea and Rennes, and the research institute of IRSTEA of Montpell ...

Joint estimation of water column parameters and seabed reflectance combining maximum likelihood and unmixing algorithm

Bathymetry and Water column constituent estimation is a challenging task for the study of coastal zones. Although most extensively used methods are based on the pixel wise inversion of semi empirical models, we have recently proposed a Maximum Likelihood (ML) method to retrieve such parameters. One limitation of the method is that the seabed reflectance is supposed to be known and homogeneous in a sample zone. We propose here to go beyond this limitation by jointly estimating the bathymetry, the concentrations of the constituents, and the reflectance spectra in an inhomogeneous seabed zone. To do so we introduce a non stationary Likelihood for the sample zone, and we also exploit a triple non-negative matrix factorization. We propose an Estimation-Unmixing (E-U) recursive algorithm to solve the problem. The water column parameters are estimated within the ML step, while the unmixing step allows to recover the bottom reflectance in each pixel. When tested on real hyper-spectral data acquired in the Quiberon peninsula on French West coast, the method leads to consistent estimated maps of bathymetry and seabed reflectance.

Multi-temporal hyperspectral data classification without explicit reflectance correction

In order to be independent from light source and atmospheric conditions, radiance values extracted from a remote hyperspectral image have to be converted into reflectance values before data processing. Several methods have been proposed in the literature but they require that the lighting/and or atmospheric conditions to be estimated. In the framework of supervised classification, we propose an approach to deal with such lighting and atmospheric temporal fluctuations without reference measurement. Assuming that materials to the surface objects to be discriminated is Lambertian, we show that the difference in lighting conditions after a log-transformation of both reflectance and radiance signals can be expressed as an additive effect. This effect remains additive after the use of a linear dimension reduction method and can be efficiently estimated in a low dimensional feature space. In the feature space, this difference in ligthing can be estimated and thus corrected by finding the translation for which the class densities obtained for each image best overlaps (using cross correlation). This novel approach was applied on a remote sensing data set over the Quiberon peninsula France. For the tested images, classification results obtained with this approach were comparable to those obtained using a classical reflectance correction technique.