Predicting rice diseases using advanced technologies at different scales: present status and future perspectives

Area of Science:

• Agricultural Science
• Computer Science
• Plant Pathology

Background:

• Accurate and rapid tracking of rice diseases is crucial for global food security.
• Emerging techniques are vital for addressing challenges in high-throughput data analysis for crop monitoring.
• Traditional methods often struggle with the complexity and scale of modern agricultural data.

Purpose of the Study:

• To review image processing techniques utilizing machine learning (ML) and deep learning (DL) for multi-scale rice disease detection.
• To summarize diverse detection approaches including genomic, physiological, and biochemical methods.
• To present the current state of optical sensing in pathogen-plant interaction phenotype analysis.

Main Methods:

• Focus on image processing techniques powered by ML and DL models.
• Review of genomic, physiological, and biochemical detection strategies.
• Analysis of contemporary optical sensing applications for pathogen-plant interactions.

Main Results:

• ML and DL models show significant promise for accurate and rapid rice disease detection.
• Integration of image processing with ML/DL offers effective solutions for complex datasets.
• Optical sensing provides insights into pathogen-plant interactions at the phenotypic level.

Conclusions:

• ML and DL are essential tools for the early detection of rice crop diseases.
• This review provides a comprehensive resource for researchers in agricultural technology and plant pathology.
• Future research should focus on refining high-throughput data analysis and model recognition for enhanced crop management.