A comprehensive image dataset of plum leaf and fruit for disease classification

Area of Science:

• Agricultural Science
• Computer Vision
• Machine Learning

Background:

• Plums (Indian jujube) are economically significant fruits, valued for nutrition and global consumption.
• Effective disease management and quality assessment in agriculture increasingly rely on advanced computational tools.
• A specialized dataset is crucial for developing machine learning models in plant pathology and agricultural monitoring.

Purpose of the Study:

• To create a comprehensive dataset of plum leaf and fruit images for machine learning applications.
• To facilitate automated disease detection and fruit quality assessment in plums.
• To support advancements in agricultural research and sustainable farming practices.

Main Methods:

• Collected 3,554 original plum images under diverse environmental conditions (Dec 2024 - Feb 2025).
• Processed and augmented the dataset to include 18,000 images.
• Categorized images into six classes: Shot Hole, Bacterial Spot, Wilted Leaf, Healthy Leaf, Unhealthy Plum, and Healthy Plum.

Main Results:

• A robust dataset comprising original, processed, and augmented plum images was established.
• The dataset is structured for machine learning-based classification of plum diseases and health status.
• It enables the development of tools for early disease identification and fruit quality monitoring.

Conclusions:

• This dataset is a foundational resource for deep learning in agriculture, specifically for plum cultivation.
• It empowers researchers to develop early disease detection systems, enhancing crop management and reducing chemical inputs.
• The dataset promotes sustainable agriculture by helping farmers minimize losses and improve produce quality.