Deep learning techniques, and in particular Convolutional Neural Networks (CNNs), have led to significant progress in image processing. Since 2016, many applications for the automatic identification of crop diseases have been developed. These applications could serve as a basis for the development of expertise assistance or automatic screening tools. Such tools could contribute to more sustainable agricultural practices and greater food production security. To assess the potential of these networks for such applications, we survey 19 studies that relied on CNNs to automatically identify crop diseases. We describe their profiles, their main implementation aspects and their performance. Our survey allows us to identify the major issues and shortcomings of works in this research area. We also provide guidelines to improve the use of CNNs in operational contexts as well as some directions for future research.

Speckle noise filtering on polarimetric SAR (PolSAR) images remains a challenging task due to the difficulty to reduce a scatterer-dependent noise while preserving the polarimetric information and the spatial information. This challenge is particularly acute on single look complex images, where little information about the scattering process can be derived from a rank-1 covariance matrix. This paper proposes to analyze and to evaluate the performances of a set of PolSAR speckle filters. The filter performances are measured by a set of ten different indicators, including relative errors on incoherent target decomposition parameters, coherences, polarimetric signatures, point target, and edge preservation. The result is a performance profile for each individual filter. The methodology consists of simulating a set of artificial PolSAR images on which the various filters will be evaluated. The image morphology is stochastic and determined by a Markov random field and the number of scattering classes is allowed to vary so that we can explore a large range of image configurations. Evaluation on real PolSAR images is also considered. Results show that filters performances need to be assessed using a complete set of indicators, including distributed scatterer parameters, radiometric parameters, and spatial information preservation.

Survey and monitoring of wildlife populations are among the key elements in nature conservation. The use of unmanned aerial vehicles and light aircrafts as aerial image acquisition systems is growing, as they are cheaper alternatives to traditional census methods. However, the manual localization and identification of species within imagery can be time‐consuming and complex. Object detection algorithms, based on convolutional neural networks (CNNs), have shown a good capacity for animal detection. Nevertheless, most of the work has focused on binary detection cases (animal vs. background). The main objective of this study is to compare three recent detection algorithms to detect and identify African mammal species based on high‐resolution aerial images. We evaluated the performance of three multi‐class CNN algorithms: Faster‐RCNN, Libra‐RCNN and RetinaNet. Six species were targeted: topis (Damaliscus lunatus jimela), buffalos (Syncerus caffer), elephants (Loxodonta africana), kobs (Kobus kob), warthogs (Phacochoerus africanus) and waterbucks (Kobus ellipsiprymnus). The best model was then applied to a case study using an independent dataset. The best model was the Libra‐RCNN, with the best mean average precision (0.80 ± 0.02), the lowest degree of interspecies confusion (3.5 ± 1.4%) and the lowest false positive per true positive ratio (1.7 ± 0.2) on the test set. This model was able to detect and correctly identify 73% of all individuals (1115), find 43 individuals of species other than those targeted and detect 84 missed individuals on our independent UAV dataset, with an average processing speed of 12 s/image. This model showed better detection performance than previous studies dealing with similar habitats. It was able to differentiate six animal species in nadir aerial images. Although limitations were observed with warthog identification and individual detection in herds, this model can save time and can perform precise surveys in open savanna.

Nowadays, satellite images are considered as one of the most relevant sources of information in the context of major disasters management. Their availability in extreme weather conditions and their ability to cover wide geographic areas make them an indispensable tool toward an effective disaster response. Among the various available sensors, Synthetic Aperture Radar (SAR) is distinguished in the context of flood management by its ability to penetrate cloud cover and its robustness to unfavourable weather conditions. This work aims at developing a new technique for flooded areas extraction from high resolution time-series SAR images. The proposed approach is mainly based on three steps: first, homogeneous regions characterizing water surfaces are extracted from each SAR image using a local texture descriptor. Then, mathematical morphology is applied to filter tiny artifacts and small homogeneous areas present in the image. And finally, spatial and radiometric information embedded in each pixel are extracted and are fused with the same pixel information but from another image to decide if the current pixel belongs to a flooded region. In order to assess the performance of the proposed algorithm, our methodology was applied to time-series images acquired before and during three different flooding events: (1) Richelieu River and lake Champlain floods, Quebec, Canada in 2011; (2) Evros River floods, Greece in 2014 and (3) Western and southwestern of Iran floods in 2016. Experiments show that our approach gives very promising results compared to existing techniques.

Water bodies extraction using satellite images is of great importance due to its utility in several applications such as land use planning, floods management, and monitoring. Among the wide range of sensors orbiting the Earth, Synthetic Aperture Radar (SAR) is a very effective tool in this context due to its robustness in the face of unfavorable weather conditions and its cloud penetration capabilities. This paper presents a novel river extraction algorithm from high-resolution SAR images mainly based on the combination of a local texture measurement and the global knowledge associated with the shape of the object of interest. A local texture measurement is first computed at every pixel of the image to extract the homogeneous surfaces contained in the image, and then a mathematical morphology operator is applied to eliminate the noise generated by speckle characterizing SAR images. Finally, the surface occupied by the object of interest is compared with the surface associated with the smallest rectangle that encloses this object in order to separate rivers from lakes in the image. The proposed approach was tested on SAR images acquired by the RADARSAT-2 satellite over numerous regions of Canada. Our experimental results demonstrate that the proposed approach is robust and effective.

In the last decade there has been a growing interest in the study of sparse representation of signals. In particular, many new multiscale image representations in a geometric space have been proposed (Curvelets, Ridgelets, Contourlets, etc.). Instead of using a fixed transformation, an alternative approach is to build a sparse dictionary from the signal itself. In the present work, we propose a novel approach for speckle noise reduction in SAR images using a sparse and redundant representation over trained dictionaries. In this approach, an adaptive dictionary composed of image patches (called atoms) is learned from the image so that it constitutes a sparse representation of the image content. This learning process, called K-SVD, is efficiently performed using an Orthogonal Matching Pursuit (OMP) and a Singular Value Decomposition (SVD). This new approach is effective in removing white additive Gaussian noise despite the fact that elements of the dictionary are learned from the noisy image, the algorithm is converging toward meaningful atoms that are already showing a reduction in noise level.