Plant disease severity estimated visually: a century of research, best practices, and opportunities for improving methods and practices to maximize accuracy

Bock, Clive H.; Chiang, Kuo-Szu; Ponte, Emerson M. Del

doi:10.1007/s40858-021-00439-z

Cited by 89 publications

(52 citation statements)

References 84 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When assessing plant disease severity visually, raters have been considered an important source of error and variability affecting accuracy (Bock et al 2021). In our study, differently from estimation, the user of pliman needs to manually select areas of the image and prepare a new image that represents the range of color variations that are specific to at least two portions: the symptomatic and the non-symptomatic area displayed in the image.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Measuring plant disease severity in R: introducing and evaluating the pliman package

Olivoto

Andrade

Ponte

2022

Trop. plant pathol.

View full text Add to dashboard Cite

Image analysis based on color thresholding is the reference method for measuring severity as percent area affected. It is deemed to produce accurate results, usually considered the "true" severity value. More than a dozen applications have been used for the task in phytopathometry studies, but none was coded in R language. Here we introduced and evaluated pliman, a suite for the analysis of plant images. In particular, we show functions for computing percent severity based on RGB information contained in image palettes prepared by the user. Six image collections, totaling 249 images, from different diseases (wheat tan spot, soybean rust, olive leaf spot, rice brown spot, bean angular spot, and Xyllela fastidiosa on tobacco) exhibiting a range of symptomatic patterns and severity were used to evaluate the agreement of pliman predictions with APS Assess, LeafDoctor and ImageJ. Three users independently prepared three image palettes (each representing leaf background, symptomatic or healthy leaf tissue) by manually inspecting and subsetting these target areas in the images. Pliman predictions by a joint palette (by joining images by the three users into one) were highly concordant (ρ c > 0.98) with measures by the other software for all but Xylella fastidiosa on Tobacco (ρ c = 0.49). The error for the latter may be due to the low contrast between symptomatic and healthy tobacco tissues. Users showed to be a source of variation in the overall concordance depending on the disease. Reduction in the image resolution (< 1 megapixel) did not impact the results. Combined with parallel processing, the use of low image resolution sped up the processing time resulting in pliman being ~170 to ~430 times faster than existing tools for disease quantification. Pliman showed great potential to produce accurate measures and accelerate studies involving plant disease severity measurements, especially for the batch processing of large sets of image collections.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Historically, and still very commonly, especially in field operations, percent severity is obtained via visual assessment, with or without aids such as standard area diagrams (SADs) (Bock et al 2021).…”

Section: Introductionmentioning

confidence: 99%

Measuring plant disease severity in R: introducing and evaluating the pliman package

Olivoto

Andrade

Ponte

2022

Trop. plant pathol.

View full text Add to dashboard Cite

show abstract

“…Visual estimation is the conventional approach to quantify the disease severity, which assign a severity value to the symptoms perceived by the human eye in the visible light range. Disease severity based on ratio scales is usually performed by manual estimation of the visual score according to the number and the area of the plants’ lesions [ 3 ]. It has been proved the most accurate tool for estimating the disease severity [ 4 ].…”

Section: Introductionmentioning

confidence: 99%

“…The accuracy for the five severity categories that were defined in this method was 86.51%. Although accurate results were reported in the above studies, dividing severity percentages into multiple categories in field trials did not make it easy to assess the effectiveness of treatments, such as fungicides [ 3 ].…”

Section: Introductionmentioning

confidence: 99%

Attention-optimized DeepLab V3 + for automatic estimation of cucumber disease severity

Zhang

et al. 2022

Plant Methods

View full text Add to dashboard Cite

Background Automatic and accurate estimation of disease severity is critical for disease management and yield loss prediction. Conventional disease severity estimation is performed using images with simple backgrounds, which is limited in practical applications. Thus, there is an urgent need to develop a method for estimating the disease severity of plants based on leaf images captured in field conditions, which is very challenging since the intensity of sunlight is constantly changing, and the image background is complicated. Results This study developed a simple and accurate image-based disease severity estimation method using an optimized neural network. A hybrid attention and transfer learning optimized semantic segmentation model was proposed to obtain the disease segmentation map. The severity was calculated by the ratio of lesion pixels to leaf pixels. The proposed method was validated using cucumber downy mildew, and powdery mildew leaves collected under natural conditions. The results showed that hybrid attention with the interaction of spatial attention and channel attention can extract fine lesion and leaf features, and transfer learning can further improve the segmentation accuracy of the model. The proposed method can accurately segment healthy leaves and lesions (MIoU = 81.23%, FWIoU = 91.89%). In addition, the severity of cucumber leaf disease was accurately estimated (R2 = 0.9578, RMSE = 1.1385). Moreover, the proposed model was compared with six different backbones and four semantic segmentation models. The results show that the proposed model outperforms the compared models under complex conditions, and can refine lesion segmentation and accurately estimate the disease severity. Conclusions The proposed method was an efficient tool for disease severity estimation in field conditions. This study can facilitate the implementation of artificial intelligence for rapid disease severity estimation and control in agriculture.

show abstract

“…Quantitative methods of phenotyping have been developed as well, including ordinal scales and sets of standard area diagrams [ 14 , 15 ], but have been used less frequently for genetic mapping research. Standard area diagrams offer increased precision over qualitative assessments, but accuracy may still be impacted by rater experience, the number of diagrams, and the quality of diagrams [ 16 ]. In the past, phenotyping severity of many plant leaf diseases have used 5-point ordinal scales, such as northern corn leaf blight [ Exserohilum turcicum (Pass.)…”

Section: Introductionmentioning

confidence: 99%

Automated, image-based disease measurement for phenotyping resistance to soybean frogeye leaf spot

McDonald

Buck

2022

Plant Methods

View full text Add to dashboard Cite

Background Frogeye leaf spot is a disease of soybean, and there are limited sources of crop genetic resistance. Accurate quantification of resistance is necessary for the discovery of novel resistance sources, which can be accelerated by using a low-cost and easy-to-use image analysis system to phenotype the disease. The objective herein was to develop an automated image analysis phenotyping pipeline to measure and count frogeye leaf spot lesions on soybean leaves with high precision and resolution while ensuring data integrity. Results The image analysis program developed measures two traits: the percent of diseased leaf area and the number of lesions on a leaf. Percent of diseased leaf area is calculated by dividing the number of diseased pixels by the total number of leaf pixels, which are segmented through a series of color space transformations and pixel value thresholding. Lesion number is determined by counting the number of objects remaining in the image when the lesions are segmented. Automated measurement of the percent of diseased leaf area deviates from the manually measured value by less than 0.05% on average. Automatic lesion counting deviates by an average of 1.6 lesions from the manually counted value. The proposed method is highly correlated with a conventional method using a 1–5 ordinal scale based on a standard area diagram. Input image compression was optimal at a resolution of 1500 × 1000 pixels. At this resolution, the image analysis method proposed can process an image in less than 10 s and is highly concordant with uncompressed images. Conclusion Image analysis provides improved resolution over conventional methods of frogeye leaf spot disease phenotyping. This method can improve the precision and resolution of phenotyping frogeye leaf spot, which can be used in genetic mapping to identify QTLs for crop genetic resistance and in breeding efforts for resistance to the disease.

show abstract

Plant disease severity estimated visually: a century of research, best practices, and opportunities for improving methods and practices to maximize accuracy

Cited by 89 publications

References 84 publications

Measuring plant disease severity in R: introducing and evaluating the pliman package

Measuring plant disease severity in R: introducing and evaluating the pliman package

Attention-optimized DeepLab V3 + for automatic estimation of cucumber disease severity

Automated, image-based disease measurement for phenotyping resistance to soybean frogeye leaf spot

Contact Info

Product

Resources

About