GuavaLeafBD: A Labeled Image Dataset of Guava Leaf Diseases for image based classification
Description
This dataset supports research in plant pathology and computer vision by providing a labeled image collection of guava leaf diseases and healthy leaves collected from guava gardens in the Chittagong Division of Bangladesh. The dataset contains 524 images representing five categories: Red Rust, Yellow Spot, Healthy, Caterpillar, and Die Back. The guava leaf images were collected from guava gardens in Kanchannagar , Chandanaish, Chittagong, Bangladesh, an area well known for guava cultivation. The images were captured using two high-end smartphones, Samsung Galaxy S25 Ultra and iPhone 15 Pro Max. The dataset contains the following class distribution: • Red Rust: 73 images • Yellow Spot: 106 images • Healthy: 101 images • Caterpillar: 207 images • Die Back: 37 images
Files
Steps to reproduce
The dataset was developed through a field-based guava leaf sample collection and image acquisition process. 1. Guava leaf samples were collected directly from guava gardens located in Kanchannagar, Chandanaish, Chittagong, Bangladesh, a region well known for guava cultivation. 2. Leaves showing visible symptoms of the targeted diseases or pest damage were selected for the corresponding disease classes. Healthy leaves without visible symptoms were collected for the Healthy class. 3. The collected leaf samples were categorized into five classes: Red Rust, Yellow Spot, Healthy, Caterpillar, and Die Back. 4. Images of the collected leaf samples were captured using two smartphones: Samsung Galaxy S25 Ultra and iPhone 15 Pro Max. 5. The captured images were organized into class-specific folders according to their corresponding labels. 6. The resulting dataset contains 524 labeled images across the five classes and can be used for image-based guava leaf disease classification and related computer vision research. No specialized laboratory reagents or instruments were required for data collection. The dataset was created using naturally occurring guava leaf samples collected from the specified guava-growing locations and smartphone-based image acquisition.
Institutions
- BGC Trust University BangladeshChittagong, Chittagong