Implementação de um Modelo Bag of Features para Classificação de Frutas

This work explores a classic technique in computer vision, the Bag of Features (BoF) model, in a fruit and vegetable classification problem. There’s an increasing trend in the use of Neural Networks and Deep Learning techniques applied to the automation of processes and systems. This work goes against this trend, examining how a simpler Machine Learning (ML) model would perform. For this, we defined two scenarios, one in a more controlled environment with differences only in light and objects positions, and another with more background clutter. We show that, although the trend is to use bigger and more complex ML models, simpler techniques continue to be relevant in certain scenarios.