Mint leaves: Dried, fresh, and spoiled dataset for condition analysis and machine learning applications

Area of Science:

• Computer Vision
• Machine Learning
• Agricultural Science

Background:

• Accurate assessment of mint leaf quality is crucial for agriculture, food preservation, and pharmaceutical industries.
• Existing methods for leaf condition analysis often lack comprehensive, diverse datasets for robust model training.
• The need for automated systems to rapidly evaluate leaf quality is increasing.

Purpose of the Study:

• To introduce a comprehensive dataset of 5,323 mint (pudina) leaf images.
• To facilitate research in condition analysis and machine learning for leaf quality assessment.
• To provide a resource for training and evaluating computer vision models for mint leaf discernment.

Main Methods:

• Collected and curated a dataset of 5,323 mint leaf images.
• Images represent various conditions: fresh, dried, and spoiled.
• Included manual annotations categorizing each image and ensured variations in lighting, background, and orientation.

Main Results:

• A diverse and annotated dataset of mint leaf images is now available.
• The dataset encompasses controlled variations to enhance model generalizability.
• Enables the training and evaluation of machine learning and computer vision algorithms.

Conclusions:

• The presented dataset is a valuable resource for advancing automated mint leaf quality assessment.
• It supports the development of reliable systems for industries requiring rapid quality evaluation.
• Encourages further research into innovative machine learning approaches for plant condition analysis.