Plant diseases are a crucial issue in agriculture. An accurate and automatic identification of leaf diseases could help to develop an early response to reduce economic losses. Recent research in plant diseases has adopted deep neural networks. However, such research has used the models as a black-box passing the labeled images through the networks. This letter presents an analysis of the network weights for the automatic recognition of soybean leaf diseases applied to images taken straight from a small and cheap unmanned aerial vehicle (UAV). To achieve high accuracy, we evaluated four deep neural network models trained with different parameters for fine-tuning (FT) and transfer learning. Data augmentation and dropout were used during the network training to avoid overfitting. Our methodology consists of using the SLIC method to segment the plant leaves in the top-view images obtained during the flight. We tested our data set created from real flight inspections in an end-to-end computer vision approach. Results strongly suggest that the FT of parameters substantially improves the identification accuracy.

Soybean has been the main Brazilian agricultural commodity, contributing substantially to the country’s trade balance. However, foliar diseases are the key factor that can undermine the soy production, usually caused by fungi, bacteria, viruses, and nematodes. This letter proposes a computer vision system to track soybean foliar diseases in the field using images captured by the low-cost unmanned aerial vehicle model DJI Phantom 3. The proposed system is based on the segmentation method Simple Linear Iterative Clustering to detect plant leaves in the images and on visual attributes to describe the features of foliar physical properties, such as color, gradient, texture, and shape. Our methodology evaluated the performance of six classifiers for different heights, including 1, 2, 4, 8, and 16 m. Experimental results showed that color and texture attributes lead to higher classification rates, achieving the precision of 98.34% for heights between 1 and 2 m, with a decay of 2% at each meter. Results indicate that our approach can support experts and farmers to monitor diseases in soybean fields.

Time-series forecasting is one of the most active research topics in artificial intelligence. It has the power to bring light to problems in several areas of knowledge, such as epidemiological studies, healthcare inference, and climate change analysis. Applications in real-world time series should consider two factors for achieving reliable predictions: modeling dynamic dependencies among multiple variables and adjusting the model's intrinsic hyperparameters. An open gap in the literature is that statistical and ensemble learning approaches systematically present lower predictive performance than deep learning methods. The existing applications consistently disregard the data sequence aspect entangled with multivariate data represented in more than one time series. Conversely, this work presents a novel neural network architecture for time-series forecasting that combines the power of graph evolution with deep recurrent learning on distinct data distributions, named after Recurrent Graph Evolution Neural Network ( ReGENN ). The idea is to infer multiple multivariate relationships between co-occurring time-series by assuming that the temporal data depends not only on inner variables and intra-temporal relationships (i.e., observations from itself) but also on outer variables and inter-temporal relationships (i.e., observations from other-selves). An extensive set of experiments was conducted comparing ReGENN with tens of ensemble methods and classical statistical ones. The results outperformed both statistical and ensemble-learning approaches, showing an improvement of 64.87 percent over the competing algorithms on the SARS-CoV-2 dataset of the renowned John Hopkins University for 188 countries simultaneously. For further validation, we tested our architecture in two other public datasets of different domains, the PhysioNet Computing in Cardiology Challenge 2012 and Brazilian Weather datasets. We also analyzed the Evolution Weights arising from the hidden layers of ReGENN to describe how the variables of the dataset interact with each other; and, as a result of looking at inter and intra-temporal relationships simultaneously, we concluded that time-series forecasting is majorly improved if paying attention to how multiple multivariate data synchronously evolve.

Corrosion identification and repair is a vital task in aircraft maintenance to ensure continued structural integrity. Regarding fuselage lap joints, typically, visual inspections are followed by non-destructive methodologies, which are time-consuming. The visual inspection of large areas suffers not only from subjectivity but also from the variable probability of corrosion detection, which is aggravated by the multiple layers used in fuselage construction. In this paper, we propose a methodology for automatic image-based corrosion detection of aircraft structures using deep neural networks. For machine learning, we use a dataset that consists of D-Sight Aircraft Inspection System (DAIS) images from different lap joints of Boeing and Airbus aircrafts. We also employ transfer learning to overcome the shortage of aircraft corrosion images. With precision of over 93%, we demonstrate that our approach detects corrosion with a precision comparable to that of trained operators, aiding to reduce the uncertainties related to operator fatigue or inadequate training. Our results indicate that our methodology can support specialists and engineers in corrosion monitoring in the aerospace industry, potentially contributing to the automation of condition-based maintenance protocols.

The occurrence of insect pest attacks in soybean fields has worried farmers around the world. Early and automatic diagnosis of insect pests number could assess the infestation level of each plantation area to optimize the applications of pesticides in the crop and, consequently, reduce production costs and environmental impact. Recent research on insect count has adopted deep neural networks. However, researches have employed models trained to count only one species of insect, using images captured in a controlled environment, quite different from a real scenario. In order to obtain high accuracy, we evaluated three models of convolutional neural networks (CNNs) with three different training strategies: 100% fine-tuning with the weights obtained from ImageNet, a complete network with the weights initialized randomly and transfer learning with the weights obtained from ImageNet. Data augmentation and dropout were used during network training to reduce overfitting and increase generalization of the model. Our approach consists in segmenting an image from the plantation with the simple linear iterative clustering (SLIC) method and classifying each superpixel segment into a pest insect class using the CNN-trained classification model. The pest insect count is obtained by adding the insects of each superpixel class identified by our computer vision system. The results indicate that the deep-learning models can be used successfully to support specialists and farmers in the insect pest management in soybean fields.

In this paper we model spatio-temporal data describing the fishing activities in the Northern Adriatic Sea over four years. We build, implement and analyze a database based on the fusion of two complementary data sources: trajectories from fishing vessels (obtained from terrestrial Automatic Identification System, or AIS, data feed) and fish catch reports (i.e., the quantity and type of fish caught) of the main fishing market of the area. We present all the phases of the database creation, starting from the raw data and proceeding through data exploration, data cleaning, trajectory reconstruction and semantic enrichment. We implement the database by using MobilityDB, an open source geospatial trajectory data management and analysis platform. Subsequently, we perform various analyses on the resulting spatio-temporal database, with the goal of mapping the fishing activities on some key species, highlighting all the interesting information and inferring new knowledge that will be useful for fishery management. Furthermore, we investigate the use of machine learning methods for predicting the Catch Per Unit Effort (CPUE), an indicator of the fishing resources exploitation in order to drive specific policy design. A variety of prediction methods, taking as input the data in the database and environmental factors such as sea temperature, waves height and Clorophill-a, are put at work in order to assess their prediction ability in this field. To the best of our knowledge, our work represents the first attempt to integrate fishing ships trajectories derived from AIS data, environmental data and catch data for spatio-temporal prediction of CPUE – a challenging task.