RESUMO
Bunk management is an important technique to minimize the variations in consumption in feedlot cattle and can be performed according to the South Dakota State University classification system. The use of information and communication technology (ICT) can help, in an objective way, in the interpretation of these measurements. We created a dataset with the objective to develop an automatic classification method of feed bunk score. In May, September and October on the 2021 and September on the 2022 we captured 1511 images in the morning on the farms, in natural lighting conditions with different angles and backgrounds and at a height of about 1.5 m from the bunk. After acquisition data, each image was classified according to its score classification. Additionally, we resized the images to 500 × 500 pixels, generated annotations files, and organized the dataset in folders. The images in this dataset can be used to train and validate a machine learning model to classify feed bunk images. This model can be used to develop an application to support bunk management.
