“Desarrollo de un prototipo para el reconocimiento de plagas en plantaciones de babaco bajo invernadero utilizando redes neuronales”.

This research project is focused on the detection of pests that occur in babaco plants, that is, a technical analysis is made by experts and through bibliographic reviews considering image processing for this project through training with convolutional neural networks whose results are reflected in a mobile interface, the same one that was designed for farmers in this type of plantation. For the development of the neural network training, the evolutionary development methodology was used, through the use of programming tools, Google Colab was used as a development platform, Python as a programming language, the Open CV library for image processing and Android Studio which was used to create the mobile interface. For training and testing of the neural network, 2500 images of pests in leaves and fruits based on photographs were used. To establish the type of plague, image processing was used in a bank of fifty photographs per plague through two training processes of the neural network, obtaining 93.75% effectiveness and 100% accuracy as results.

• Автор: Gamboa Murillo, Patricio Alejandro (author)
• Інші автори: López Miniguano, Isaac Samuel (author)
• Формат: bachelorThesis
• Мова: spa
• Опубліковано: 2021
• Предмети: PROCESAMIENTO; PLAGAS; GOOGLE COLAB; OPEN CV; SISTEMAS
• Онлайн доступ: http://repositorio.utc.edu.ec/handle/27000/8625