Comprehensive guava fruit data set: Digital and thermal images for analysis and classification.

Area of Science:

• Agricultural Science
• Computer Vision
• Image Processing

Background:

• Guava (Psidium guajava) production is threatened by declining yields due to infections and diseases.
• Early detection of guava diseases is crucial for mitigating economic losses and ensuring sustainable farming practices.

Purpose of the Study:

• To present a comprehensive dataset for the early identification of guava diseases.
• To facilitate the development of expert systems for prompt disease diagnosis by farmers.

Main Methods:

• The dataset comprises digital and thermal images of guava fruits exhibiting healthy, damaged, and diseased conditions (wilt, Anthracnose, canker, rot).
• Images are categorized by maturity level and captured under various drop heights, incorporating details on damage induction, storage, and capture conditions.
• Thermal images were acquired using controlled hot air application.

Main Results:

• The dataset provides a rich resource for training machine learning models for guava disease identification.
• Enables the development of advanced image processing techniques for distinguishing between healthy and diseased guava fruits.
• Facilitates differentiation of various disease types based on visual and thermal characteristics.

Conclusions:

• The developed dataset is vital for advancing automated disease detection systems in guava cultivation.
• This technological approach supports sustainable agriculture by enabling timely interventions against fruit diseases.
• Empowers farmers with tools for early disease identification, thereby improving crop management and reducing economic impact.