February 20, 2025 | Data in Brief |
This study conducted by Daffodil International University, Bangladesh developed an image dataset to support automated detection of diseases affecting guava (Psidium guajava), an economically and nutritionally important tropical fruit widely cultivated in Bangladesh. Guava production is often threatened by various fruit and leaf diseases that can significantly reduce yield and fruit quality. Early detection is therefore essential for effective management and for minimizing economic losses.
To support the development of automated disease detection systems, the researchers compiled a dataset consisting of 3,432 real images of guava fruits and leaves collected from different locations in Bangladesh. The dataset includes both healthy samples and those affected by common diseases such as anthracnose, scab, styler end rot, canker, rust, and leaf spot. To enhance its usability for machine learning applications, the dataset was expanded through data augmentation techniques, resulting in a total of 20,344 images.
The dataset is designed to facilitate research in machine learning and computer vision for agricultural disease diagnosis. By providing a structured and labeled image collection, the study enables the development and testing of automated systems capable of detecting guava diseases at early stages. Such tools could support farmers and researchers in monitoring crop health, reducing yield losses, and improving disease management in guava cultivation.
