The coffee plant is one of the main crops grown in Brazil. However, strategies to estimate its yield are questionable given the characteristics of this crop; in this context, robust techniques, such as those based on machine learning, may be an alternative. Thus, the aim of the present study was to estimate the yield of a coffee crop using multispectral images and machine learning algorithms. Yield data from a same study area in 2017, 2018 and 2019, Sentinel 2 images, Random Forest (RF) algorithms, Support Vector Machine (SVM), Neural Network (NN) and Linear Regression (LR) were used. Statistical analysis was performed to assess the absolute Pearson correlation and coefficient of determination values. The Sentinel 2 satellite images proved to be favorable in estimating coffee yield. Despite the low spatial resolution in estimating agricultural variables below the canopy, the presence of specific bands such as the red edge, mid infrared and the derived vegetation indices, act as a countermeasure. The results show that the blue band and green normalized difference vegetation index (GNDVI) exhibit greater correlation with yield. The NN algorithm performed best and was capable of estimating yield with 23% RMSE, 20% MAPE and R² 0.82 using 85% of the training and 15% of the validation data of the algorithm. The NN algorithm was also more accurate (27% RMSE) in predicting yield.

Image-based spectral models assist in estimating the yield of maize. During the vegetative and reproductive phenological phases, the corn crop undergoes changes caused by biotic and abiotic stresses. These variations can be quantified using spectral models, which are tools that help producers to manage crops. However, defining the correct time to obtain these images remains a challenge. In this study, the possibility to estimate corn yield using multispectral images is hypothesized, while considering the optimal timing for detecting the differences caused by various phenological stages. Thus, the main objective of this work was to define the ideal phenological stage for taking multispectral images to estimate corn yield. Multispectral bands and vegetation indices derived from the Planet satellite were considered as predictor variables for the input data of the models. We used root mean square error percentage and mean absolute percentage error to evaluate the accuracy and trend of the yield estimates. The reproductive phenological phase R2 was found to be optimal for determining the spectral models based on the images, which obtained the best root mean square error percentage of 9.17% and the second-best mean absolute percentage error of 7.07%. Here, we demonstrate that it is possible to estimate yield in a corn plantation in a stage before the harvest through Planet multispectral satellite images.

In Brazil there are six well-defined biomes and the Caatinga represents 9.92% of the total area. This biome is exclusively Brazilian and very rich in biodiversity. Because it has low resistance to human interference is necessary to know the important factors in monitoring the biome. Vegetation coverage and climate are two of these factors, as they indicate the intensity of human activity and the wear caused with time. Normalized Difference Vegetation Index (NDVI) has been recently explored in the description of tree phenology. In this study we sought to associate NDVI/Landsat values to a climatic variable (phenomena El-Nino and La-Nina - NINONINA) in order to describe the phenological behavior of the Caatinga in the Parque Nacional da Serra das Capivaras in the state of Piauí/Brazil. A time series analysis was carried out, describing the intrinsic parameters of the series (Seasonality and Trends), the forecast of NDVI values using SARIMA model and co-integration between NDVI and NINONINA series. The results showed that the NDVI series presents seasonally, but does not exhibit a trend. The forecasting process presenting relatively low error at a 95% confidence interval. Finally, it was observed that the NDVI series is co-integrated with the NINONINA series.

A Spodoptera frugiperda (Smith) (lagarta-do-cartucho) é a principal praga do milho, com a intensificação da agricultura, os cultivos sucessivos possibilitam maior infestação pela praga. Isso levou ao surgimento de populações resistentes a inseticidas e culturas transgênicas. Pesquisas de campo para iniciar tratamentos com inseticidas são demoradas e exaustivas. Pensando em agilidade e qualidade, neste trabalho foi utilizado uma aeronave remotamente pilotada (ARP) equipado com uma câmera RGB e uma câmera MAPPIR 3 para capturar imagens de uma lavoura de milho, com o objetivo de estimar o índice de área foliar (IAF) de um talhão infestado por S. frugiperda. Durante o ciclo da cultura do milho foram realizadas várias avaliações: determinação do índice de área foliar (IAF), severidade do ataque da praga e, voos para aquisição das imagens. Modelos radiométricos para estimativa do IAF foram obtidos a partir de modelos de regressão linear compostos pelas bandas que melhor correlacionaram com os parâmetros medidos. Os resultados obtidos demonstraram eficiência e maior precisão na estimativa do IAF para o modelo radiométrico composto pela a banda do infravermelho próximo na câmera MAPPIR 3. Nesta ocasião, o RMSE calculado foi de 885,0714 cm². Use of images for attack detection of Spodoptera frugiperda in corn under function of loss of foliar area A B S T R A C TSpodoptera frugiperda (Smith) (cartridge-caterpillar) is the main pest of maize, with the intensification of agriculture, the successive crops allow greater infestation by the pest. This has led to the emergence of populations resistant to insecticides and transgenic crops. Field research to begin treatments with insecticides is time consuming and exhaustive. Thus, thinking of agility and quality, in this work was used a remotely piloted aircraft (ARP) equipped with an RGB camera and a MAPPIR 3 camera to capture images of a maize crop with the objective of estimating the leaf area index (LAI) of a field infested by S. frugiperda. The radiometric models for the estimation of LAI were obtained from linear regression models composed by the bands that best correlated with the measured parameters. The obtained results demonstrated efficiency and greater accuracy in the estimation of the LAI for the radiometric model composed by the near infrared band (IVP) in the MAPPIR 3 camera. On this occasion, the RMSE calculated was 885.0714 cm².Keywords: Corn; Cartridge Caterpillar; Remotely Piloted Aircraft.

The application of biological products in agricultural crops has become increasingly prominent. The growth-promoting bacterium Azospirillum brasilense has been used as an alternative to promote greater yield in maize crops. In the context of precision agriculture, interpreting geospatial data has allowed for monitoring the effect of the application of products that increase the yield of corn crops. The objective of this work was to evaluate the potential of Kriging techniques and spectral models through images in estimating the gain in yield of maize crop after applying A. brasilense. Analyses were carried out in two commercial areas treated with A. brasilense. The results revealed that models of yield prediction by Kriging with a high volume of training data estimated the yield gain with a root-mean-square error deviation (RMSE%), mean absolute percentage error (MAPE%), and R² to be 6.67, 5.42, and 0.88, respectively. For spectral models with a low volume of training data, yield gain was estimated with RMSE%, MAPE%, and R² to be 9.3, 7.71, and 0.80, respectively. The results demonstrate the potential to map the spatial distribution of productivity gains in corn crops following the application of A. brasilense.

Neste trabalho, avaliou-se o potencial do uso de imagens multiespectrais livres para identificar a infecção por nematoides em uma lavoura cafeeira. Adotou-se uma área de estudo com ocorrência infecção por nematoides (N. paranaenses) na cultura cafeeira. Em segundo lugar, analisou-se a variação da resposta espectral do café infectados e sadio, em bandas do satélite Sentinel 2. Como hipótese, esperava-se que plantas saudáveisrespondessem diferentemente das sadias. Assim, detectamos bandas cuja variação na refletância poderia auxiliar no processo de classificação de imagens. Assim, detectou-se essas variações, observando as bandas (vermelha, IVP, vermelho limítrofe 3-4). Posteriormente, fizemos a classificação da imagem (rede neural) aplicando diferentes combinações de imagens, considerando os resultados da etapa anterior, além de uma imagem representando o índice NDVI. A combinação das bandas Red, NIR e NDVI como entrada de classificação proporcionou o melhor resultado quando comparado com as outras combinações. Essa combinação permitiu detectar áreas de infecção por nematoides e realizar a classificação da imagem com 97,91% de acurácia. Portanto, obteve-se o potencial positivo do uso de imagens livres do Sentinel 2 para identificar infecções por nematoides em lavouras cafeeiras. Este é um resultado notável, uma vez que se produziu uma solução inovadora, de baixo custo e confiante.