CoLeaf-DB: Peruvian coffee leaf images dataset for coffee leaf nutritional deficiencies detection and classification

Area of Science:

• Agricultural Science
• Computer Vision
• Plant Pathology

Background:

• Nutritional deficiencies significantly impact coffee plant yield and quality.
• Accurate identification of these deficiencies is crucial for effective crop management.
• Existing datasets for coffee leaf analysis are limited, particularly for diverse nutritional deficiencies.

Purpose of the Study:

• To introduce and describe the CoLeaf dataset, a novel collection of coffee leaf images.
• To provide a resource for training and validating deep learning models for nutritional deficiency classification in coffee plants.
• To facilitate research in automated plant health monitoring for coffee cultivation.

Main Methods:

• Identification of nutritional deficiencies by agronomists in coffee plantations in Jaén, Peru.
• Design of a controlled environment for image capture.
• Acquisition of 1006 high-resolution digital images of coffee leaves (varieties: CATIMOR, CATURRA, BORBON).

Main Results:

• The CoLeaf dataset comprises 1006 images categorized by specific nutritional deficiencies (Boron, Iron, Potassium, Calcium, Magnesium, Manganese, Nitrogen, and others).
• The dataset is structured to support the training and validation of deep learning algorithms.
• Images were captured under controlled conditions to ensure consistency and quality.

Conclusions:

• The CoLeaf dataset represents a valuable, publicly available resource for advancing automated coffee plant nutritional status assessment.
• This dataset can accelerate the development of AI-driven tools for precision agriculture in coffee farming.
• Further research can leverage CoLeaf for improved disease and deficiency detection models.