Banana and Guava dataset for machine learning and deep learning-based quality classification

In the field of agriculture, the identification and classification of fruits have become necessary for sustainable growth in horticulture sectors. Recently, different types of advanced techniques have been used for the identification and classification of fruits. Machine learning is an emerging technology with robust application in different sectors that can be used for the classification of fruits. The basic requirement for the machine learning techniques is the availability of the dataset to develop a robust machine learning model. But the limited availability of the data set is one of the major challenges in this sector. The classification of fruits using non-destructive methods can be executed through the availability of a comprehensive dataset. The rapid and precise classification of the fruit according to their quality and maturity is the desired need of the fruit storage, processing, and export industries. Therefore, this article provides a comprehensive dataset for fruits (banana and guava) and their classification according to their quality classes. The dataset was acquired using a Redmi Note 10-Pro mobile camera in natural sunlight at different angles for different classes in JPG format. The acquired dataset was classified into three different categories, Class A, Class B, and Defect, based on their physiological changes. The obtained datasets can be used for the development of different machine learning models for the quality classification of banana and guava.