Deep Learning for Tomato Disease Detection with YOLOv8

In plain language

Tomato farming faces substantial variations in crop yield because of plant diseases, making automated detection an important area of development. An automated system using the YOLOv8 deep learning algorithm was evaluated for detecting tomato diseases on an augmented dataset obtained from Roboflow. The model reached an overall detection accuracy of 66.67 per cent. Performance varied across different disease classes, revealing ongoing difficulties in distinguishing between specific conditions. Addressing these performance differences will require future work on balancing training data, testing alternative model architectures, and utilising disease-specific evaluation metrics. Establishing these computational methods provides an initial basis for developing automated disease monitoring tools designed to support agricultural yields, crop quality, and sustainable farming practices.

Key takeaways

A YOLOv8 deep learning model was applied to an augmented Roboflow dataset for automated tomato disease identification.
The detection model achieved an overall accuracy rate of 66.67 per cent.
Accuracy varied across specific disease classes, showing persistent difficulties in differentiating particular plant conditions.
Further technical improvements will require data balancing, alternative neural network architectures, and disease-specific performance metrics.

Why it matters

Plant diseases can drastically lower tomato yields, threatening agricultural productivity and food supply. Automated detection using deep learning offers a pathway towards timely intervention. Demonstrating how algorithms such as YOLOv8 perform on disease datasets highlights both the current capabilities and the technical bottlenecks, helping developers refine computer vision tools to protect crops and support sustainable agriculture.

Commercialisation angle

This work points towards automated monitoring software for agricultural producers seeking early detection of crop diseases. However, the technology is at an early research stage, having achieved 66.67 per cent overall accuracy with uneven classification across disease types. Practical field deployment by growers or agritech providers will require further development, particularly around dataset balancing, refined architectures, and targeted performance metrics.

AI-generated from the published abstract. Always read the original work before citing.

Abstract

Tomato production plays a crucial role in Saudi Arabia, with significant yield variations due to factors such as diseases. While automation offers promising solutions, accurate disease detection remains a challenge. This study proposes a deep learning approach based on the YOLOv8 algorithm for automated tomato disease detection. Augmenting an existing Roboflow dataset, the model achieved an overall accuracy of 66.67%. However, class-specific performance varies, highlighting challenges in differentiating certain diseases. Further research is suggested, focusing on data balancing, exploring alternative architectures, and adopting disease-specific metrics. This work lays the foundation for a robust disease detection system to improve crop yields, quality, and sustainable agriculture in Saudi Arabia.

Research topics

Smart Agriculture and AI
Plant Disease Management Techniques
Date Palm Research Studies

Sustainable Development Goals

2Zero Hunger

Discussion

Discuss this research

Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.

No discussion yet. Open the first thread.

New to MARATTO™? Create a free account.