From visual estimates to fully automated sensor-based measurements of plant disease severity: status and challenges for improving accuracy

Abstract

The severity of plant diseases, traditionally the proportion of the plant tissue exhibiting symptoms, is a key quantitative variable to know for many diseases and is prone to error. Good quality disease severity data should be accurate (close to the true value). Earliest quantification of disease severity was by visual estimates. Sensor-based image analysis including visible spectrum and hyperspectral and multispectral sensors are established technologies that promise to substitute, or complement visual ratings. Indeed, these technologies have measured disease severity accurately under controlled conditions but are yet to demonstrate their full potential for accurate measurement under field conditions. Sensor technology is advancing rapidly, and artificial intelligence may help overcome issues for automating severity measurement under hyper-variable field conditions. The adoption of appropriate scales, training, instruction and aids (standard area diagrams) has contributed to improved accuracy of visual estimates. The apogee of accuracy for visual estimation is likely being approached, and any remaining increases in accuracy are likely to be small. Due to automation and rapidity, sensor-based measurement offers potential advantages compared with visual estimates, but the latter will remain important for years to come. Mobile, automated sensor-based systems will become increasingly common in controlled conditions and, eventually, in the field for measuring plant disease severity for the purpose of research and decision making.

Severity assessment in the Nicotiana tabacum-Xylella fastidiosa subsp. pauca pathosystem: design and interlaboratory validation of a standard area diagram set

“…Previous studies have clearly demonstrated that overestimation is a more common issue when symptoms are comprised of small and numerous lesions (Sherwood et al 1983;Bock et al 2008;Bock et al 2020). This pattern was quite evident in this study even using four reference diagrams from minimum to 15% severity.…”

“…For foliar diseases that cause necrosis, associated or not with chlorosis, a percent severity value (a ratio variable), as opposed to descriptive or scale-based scores, is usually the most informative variable for pathogen biology and disease epidemiology studies as well as evaluation of control methods (Bock et al 2016). Thus far, the most used time-effective method to obtain direct estimates of percent severity, depending on the objective and scale, is through perception of colors in the visible spectrum by human raters-these should be capable of identifying the target symptoms and assigning percent values as close as possible to the "actual" severity value (Bock et al 2020). The task is subjective and the accuracy (closeness to the actual value) of the visual estimates varies from person to person (Bock et al 2016).…”

Digital imaging outperforms traditional scoring methods of spittlebug tolerance inUrochloa humidicolahybrids

“…Our results showed increasing values of CCC and r between visual scores from an expert and inexpert evaluator with each passing week until the end of the evaluation period (data not shown). This suggests that the inexpert evaluator got better with time in the visual scoring of plant damage, as shown elsewhere (Bock et al 2016;Bock et al 2020). Albeit the improvement gained by the inexpert evaluator, this one was not able to distinguish percentages of damage below 20% intervals, whereas the expert evaluator could do it at 10% intervals (data not shown).…”

“…The success of training of a new evaluator is dependent to inherent characteristics of such individual (e.g., previous knowledge of the plants; this case), which likely affects the accuracy of any evaluation. Bock et al (2020) recently reviewed inter-rater variability and success of training among drawbacks of visual estimates of plant damage. Albeit estimates of damage from the expert and inexpert evaluators were similar, measures of data variability (i.e., standard deviation) from the inexpert evaluator were greater than those from the expert evaluator.…”

RUST: A Robust, User-Friendly Script Tool for Rapid Measurement of Rust Disease on Cereal Leaves

“…In recent years, there has been considerable interest and progress in the development of digital image analysis systems for the assessment of plant diseases [ 20 , 21 ]. Through digital image analysis, it is possible to streamline processes and analyze information [ 21 , 22 ].…”