Browsing by Author "Silva, Mateus Coelho"

Now showing 1 - 8 of 8
  • No Thumbnail Available
    Item
    Bringing deep learning to the fields and forests : leaf reconstruction and shape estimation.
    (2022) Silva, Mateus Coelho; Bianchi, Andrea Gomes Campos; Ribeiro, Sérvio Pontes; Oliveira, Ricardo Augusto Rabelo
    One of the indicators of ecosystem health is leaf health. Among the leading indicators studied in leaves, herbivory and dis- ease presence are relevant indicators of ecosystem behavior. Several methods in the literature study leaf damage estimation processes. Most of the previous studies display the usage of methods that display limited generalism. In previous work, we displayed the possibility of using conditional GANs to estimate the leaf damage. Our results displayed that this approach increases the generalism of the solution by seeking to reconstruct the original leaf shape. In this paper, we present a deeper discussion on the results and a previous discussion on how to transport this method to the forests. We present the whole method used for approximating the leaf shape, assessing further results that display the method’s robustness. Finally, we also show preliminary methods that can be used to embed this method in edge computing hardware.
  • No Thumbnail Available
    Item
    Deep learning approach at the edge to detect iron ore type.
    (2022) Klippel, Emerson; Bianchi, Andrea Gomes Campos; Silva, Saul Emanuel Delabrida; Silva, Mateus Coelho; Garrocho, Charles Tim Batista; Moreira, Vinicius da Silva; Oliveira, Ricardo Augusto Rabelo
    There is a constant risk of iron ore collapsing during its transfer between processing stages in beneficiation plants. Existing instrumentation is not only expensive but also complex and challenging to maintain. In this research, we propose using edge artificial intelligence for early detection of landslide risk based on images of iron ore transported on conveyor belts. During this work, we defined the device edge and the deep neural network model. Then, we built a prototype will to collect images that will be used for training the model. This model will be compressed for use in the device edge. This same prototype will be used for field tests of the model under operational conditions. In building the prototype, a real-time clock was used to ensure the synchronization of image records with the plant’s process information, ensuring the correct classification of images by the process specialist. The results obtained in the field tests of the prototype with an accuracy of 91% and a recall of 96% indicate the feasibility of using deep learning at the edge to detect the type of iron ore and prevent its risk of avalanche.
  • No Thumbnail Available
    Item
    Desenvolvimento de algoritmos de IA para dispositivos vestíveis utilizando computação de borda.
    (2023) Silva, Jonathan Cristovão Ferreira da; Oliveira, Ricardo Augusto Rabelo; Silva, Mateus Coelho; Oliveira, Ricardo Augusto Rabelo; Silva, Saul Emanuel Delabrida; Nacif, José Augusto Miranda; Amorim, Vicente José Peixoto de
    Os dispositivos vestíveis estão cada vez mais presentes em nossas vidas. Além disso, os algoritmos de inteligência artificial vêm se tornando essenciais para com- por estes dispositivos. Como os dispositivos vestíveis são restritos de recursos, tec- nologias que exigem grande capacidade computacional podem ser inviáveis para aplicações neste contexto, principalmente quando se trata da computação de borda. Visto isso, o trabalho propõe o desenvolvimento de algoritmos de inteligência arti- ficial para integração nestes dispositivos com o processamento dos dados na borda, sem utilizar recursos em nuvem. Esta proposta é validada com base em dois estudos de casos. O primeiro estudo de caso é a aplicação de técnicas de Machine Learning e Deep Learning na agricultura, com o objetivo de desenvolver um capacete inte- ligente para realizar inspeção de doençãs em laranjas. No segundo estudo de caso ́e desenvolvida uma nova solução vestível para o reconhecimento de atividade de caminhada. Com o auxílio de três algoritmos de IA, este estudo de caso apresentou novas perspectivas para autoavaliação do usuário a partir dos dados coletados na atividade realizada. Dessa maneira, esse trabalho apresenta uma análise de aspec- tos do desenvolvimento de algoritmos de IA para integração em dois dispositivos vestíveis através da computação de borda.
  • No Thumbnail Available
    Item
    Edge computing smart healthcare cooperative architecture for COVID-19 medical facilities.
    (2022) Silva, Mateus Coelho; Bianchi, Andrea Gomes Campos; Ribeiro, Sérvio Pontes; Silva, Jorge Sá; Oliveira, Ricardo Augusto Rabelo
    Intelligent healthcare systems are a topic of interest in recent approaches due to novel possibilities created from edge hardware and software development. In 2020, the COVID-19 pandemic displayed the urge to speed up technological systems development to aid medical facilities. In this context, solutions must enhance the experience of both patients and healthcare professionals. Thus, we propose a novel cooperative architecture to improve healthcare facilities involved in pandemic control. On the one hand, this solution helps a faster recognition and link to the patients’ data using reality augmentation resources. On the other hand, it helps monitor the conditions of medical pro- fessionals working in the facility and exposed to contamination danger.
  • No Thumbnail Available
    Item
    Real-time systems implications in the blockchain-based vertical integration of industry 4.0.
    (2020) Silva, Mateus Coelho; Ferreira, Célio Márcio Soares; Cavalcanti, Carlos Frederico Marcelo da Cunha; Oliveira, Ricardo Augusto Rabelo
  • No Thumbnail Available
    Item
    Using mobile edge AI to detect and map diseases in citrus orchards.
    (2023) Silva, Jonathan Cristovão Ferreira da; Silva, Mateus Coelho; Luz, Eduardo José da Silva; Silva, Saul Emanuel Delabrida; Oliveira, Ricardo Augusto Rabelo
    Deep Learning models have presented promising results when applied to Agriculture 4.0. Among other applications, these models can be used in disease detection and fruit counting. Deep Learning models usually have many layers in the architecture and millions of parameters. This aspect hinders the use of Deep Learning on mobile devices as they require a large amount of processing power for inference. In addition, the lack of high-quality Internet connectivity in the field impedes the usage of cloud computing, pushing the processing towards edge devices. This work describes the proposal of an edge AI application to detect and map diseases in citrus orchards. The proposed system has low computational demand, enabling the use of low-footprint models for both detection and classification tasks. We initially compared AI algorithms to detect fruits on trees. Specifically, we analyzed and compared YOLO and Faster R-CNN. Then, we studied lean AI models to perform the classification task. In this context, we tested and compared the performance of MobileNetV2, EfficientNetV2-B0, and NASNet-Mobile. In the detection task, YOLO and Faster R-CNN had similar AI performance metrics, but YOLO was significantly faster. In the image classification task, MobileNetMobileV2 and EfficientNetV2-B0 obtained an accuracy of 100%, while NASNet-Mobile had a 98% performance. As for the timing performance, MobileNetV2 and EfficientNetV2-B0 were the best candidates, while NASNet-Mobile was significantly worse. Furthermore, MobileNetV2 had a 10% better performance than EfficientNetV2-B0. Finally, we provide a method to evaluate the results from these algorithms towards describing the disease spread using statistical parametric models and a genetic algorithm to perform the parameters’ regression. With these results, we validated the proposed pipeline, enabling the usage of adequate AI models to develop a mobile edge AI solution.
  • No Thumbnail Available
    Item
    Uso de equipamento vestível para captura do ambiente florestal : estudo de caso : contagem de folhas de dossel.
    (2019) Silva, Mateus Coelho; Oliveira, Ricardo Augusto Rabelo; Ribeiro, Sérvio Pontes; Oliveira, Ricardo Augusto Rabelo; Ribeiro, Sérvio Pontes; Nacif, José Augusto Miranda; Silva, Saul Emanuel Delabrida
    O crescimento do uso de dispositivos vestíveis com computadores embarcados é um aspecto observado tanto em aplicações comerciais quanto no ambiente acadêmico. Apesar disso, ainda há poucos registros de pesquisa e cooperação no sentido de desenvolver equipamentos vestíveis para pesquisa em campo. Uma das áreas onde esse tipo de tecnologia pode ser aplicado para melhorar o processo de pesquisa ´e a ecologia. Especialmente no estudo de dosséis florestais, as restrições de segurança e processos manuais apresentam um potencial de aplicação e melhoria através dispositivos vestíveis. Dessa maneira, esse trabalho apresenta uma análise de aspectos do desenvolvimento de protótipos para o uso de equipamento vestível na realização de tarefas de captura do ambiente florestal.
  • No Thumbnail Available
    Item
    Wearable edge AI applications for ecological environments.
    (2021) Silva, Mateus Coelho; Silva, Jonathan Cristovão Ferreira da; Silva, Saul Emanuel Delabrida; Bianchi, Andrea Gomes Campos; Ribeiro, Sérvio Pontes; Silva, Jorge Sá; Oliveira, Ricardo Augusto Rabelo
    Ecological environments research helps to assess the impacts on forests and managing forests. The usage of novel software and hardware technologies enforces the solution of tasks related to this problem. In addition, the lack of connectivity for large data throughput raises the demand for edge-computing-based solutions towards this goal. Therefore, in this work, we evaluate the opportunity of using a Wearable edge AI concept in a forest environment. For this matter, we propose a new approach to the hardware/software co-design process. We also address the possibility of creating wearable edge AI, where the wireless personal and body area networks are platforms for building applications using edge AI. Finally, we evaluate a case study to test the possibility of performing an edge AI task in a wearable-based environment. Thus, in this work, we evaluate the system to achieve the desired task, the hardware resource and performance, and the network latency associated with each part of the process. Through this work, we validated both the design pattern review and case study. In the case study, the developed algorithms could classify diseased leaves with a circa 90% accuracy with the proposed technique in the field. This results can be reviewed in the laboratory with more modern models that reached up to 96% global accuracy. The system could also perform the desired tasks with a quality factor of 0.95, considering the usage of three devices. Finally, it detected a disease epicenter with an offset of circa 0.5 m in a 6 m × 6 m × 12 m space. These results enforce the usage of the proposed methods in the targeted environment and the proposed changes in the co-design pattern.