👨🏽💻 College graduation project aimed at creating an object detection custom model in TensorFlow that performs recognition and counting of coffee plants in a coffee plantation.
PythonGPL-3.0
Machine learning model to detection and counting of coffee feet (Coffea sp.) by
video analysis
The development of the research was based on the following stages
for its realization: preparation of the images, configuration of the pre-trained models for
customized training to the coffee trees, training of the model, testing of model with images
captured by the authors at the “Shuji Nishimura” Technology College in the city of Pompeia,
implementation of the model for detection in coffees trees in video format and development of
the cumulative count of tree. The result generated was a machine learning model trained to
detect coffee trees by video and cumulatively count each unit in the planting row. However, the
adoption of this type of technology can contribute positively to the analysis of productivity by
planted area and work together with specialists and companies that want to have this
technology in their equipment.
