Guava Leaf Disease Image Dataset for Classification and Detection of Common Pathologies from Bangladesh

Published: 5 November 2025| Version 2 | DOI: 10.17632/46spm4z4d3.2
Contributors:
Md Naimul Hasan, Md Fahimur Rahman

Description

This dataset contains high-quality images of guava (Psidium guajava) leaves, categorized into six classes — one healthy class and five disease-affected classes: Spiraling Whitefly, Red Rust, Citrus Canker, Anthracnose, and Algal Spot. The images were collected from guava plants in Dhaka, Bangladesh, under natural field conditions. Each image captures the distinct visual characteristics of guava leaves, later preprocessed by removing the background and placing each leaf on a plain white background to ensure uniformity. All images were resized to 512×512 pixels to maintain consistency for computer vision and deep learning applications.

Files

Institutions

  • Daffodil International University

Categories

Botany, Fruit, Leaf Area, Guava, Agriculture

Licence