A dataset of pomegranate growth stages for machine learning-based monitoring and analysis

Machine learning and deep learning have grown very rapidly in recent years and are widely used in agriculture. Neat and clean datasets are a major requirement for building accurate and robust machine learning models and minimizing misclassification in real-time environments. To achieve this goal, we created a dataset of images of pomegranate growth stages. These images of pomegranate growth stages were taken from May to September from an orchard inside the Henan Institute of Science and Technology in China. The dataset contains 5857 images of pomegranates at different growth stages, which are labeled and classified into five periods: bud, flower, early-fruit, mid-growth and ripe. The dataset consists of four folders, which respectively store the images, two formats of annotation files, and the record files for the division of training, validation, and test sets. The authors have confirmed the usability of this dataset through previous research. The dataset may help researchers develop computer applications using machine learning and computer vision algorithms.