Current State of Hyperspectral Remote Sensing for Early Plant Disease Detection: A Review

Author: mycolabadmin
2022-01-19
View Source

Summary

This review examines how hyperspectral imaging technology can be used to detect plant diseases before visible symptoms appear. This technology works by capturing detailed light reflection data from plants that can reveal early signs of disease stress. The research has important implications for agriculture and food security. Key impacts on everyday life: – Earlier detection of crop diseases could reduce pesticide use and improve food safety – More efficient crop monitoring could lead to lower food production costs – Better disease management could help prevent crop losses and improve food security – Environmental benefits from more targeted and reduced pesticide applications – Potential for automated disease detection systems in both commercial farming and home gardening

Background

Hyperspectral remote sensing (HRS) has emerged as a promising tool for early detection of plant diseases by providing detailed spectral data that can identify subtle changes in plant health before visible symptoms appear. Recent developments in miniature hyperspectral sensors and platforms have expanded the potential applications for monitoring crop health and disease spread. However, there are still methodological gaps in experimental approaches and interpretation of spectral data for reliable early disease detection.

Objective

The main objectives were to: 1) Analyze and prove the possibility of early plant disease detection using hyperspectral remote sensing across different crops, 2) Verify if spectral reflectance bands coincide for same diseases and plants, 3) Systematize current research in HRS-based plant disease detection, and 4) Identify key gaps and challenges in the field. The review focused on four major crop types – oil palm, citrus, Solanaceae family plants, and wheat – as representative cases.

Results

The review found that early disease detection using HRS is possible with accuracies ranging from 60-95%. However, there was significant variation in important spectral bands identified across studies, even for the same crop-disease combinations. Key gaps identified include: lack of standardized experimental methods, insufficient consideration of plant phenotype/genotype effects, inadequate accounting of abiotic stress factors, and limited understanding of biochemical basis for spectral changes. Hyperspectral cameras showed better results for early detection compared to spectrometers alone.

Conclusion

While HRS shows strong potential for early plant disease detection, several methodological improvements are needed for reliable implementation. These include: standardized experimental protocols considering plant physiology and environmental factors, better understanding of biochemical bases for spectral changes, and improved data normalization approaches. Future research should focus on controlled laboratory studies before field applications and involve multidisciplinary collaboration between plant scientists and remote sensing experts.

Published in:Sensors (Basel),
Study Type:Review,
Source: 10.3390/s22030757