CS-D: A dataset for disease in tea leaf (Camellia sinensis) from Assam

Area of Science:

• Agricultural Science
• Plant Pathology
• Computer Vision

Background:

• Tea quality is significantly impacted by diseases affecting tea leaves.
• Accurate identification of tea leaf diseases is essential for maintaining crop health and yield.
• Existing datasets may not adequately represent the diversity of tea leaf pathologies.

Purpose of the Study:

• To present a comprehensive dataset of healthy and diseased tea leaves for research.
• To facilitate the development of machine learning models for automated tea leaf disease detection.
• To support advancements in agricultural technology for tea cultivation.

Main Methods:

• Collected images of tea leaves from seven distinct locations in Assam over seven months.
• Utilized three smartphone devices with varying pixel resolutions for data acquisition.
• Preprocessed the raw image data through resizing, augmentation, and cleaning techniques.

Main Results:

• The dataset includes five specific classes of diseased tea leaves: Tea Mosquito Bug, Red Spider Mite, Leaf Red Rust, Blister Blight, and Brown Blight.
• The dataset comprises both healthy and diseased tea leaf samples, offering a balanced view for model training.
• The data underwent rigorous cleaning and augmentation to enhance its utility for machine learning applications.

Conclusions:

• The proposed dataset serves as a valuable resource for researchers in plant pathology and computer vision.
• This dataset will enable the creation of more robust and accurate AI models for identifying tea leaf diseases.
• The availability of this dataset is expected to accelerate innovation in precision agriculture for the tea industry.