Accurate severity assessment of wheat stripe rust caused by Puccinia striiformis f. sp. tritici is of great significance for phenotypic determination, prediction, and control of the disease. To achieve accurate severity assessment of the disease based on the actual percentages of lesion areas in the areas of the corresponding whole diseased leaves, two new methods were proposed for severity assessment of the disease. In the Adobe Photoshop 2022 software, the acquired images of single diseased leaves of each severity class of the disease were manually segmented, and the numbers of the leaf region pixels and lesion pixels of each diseased leaf were obtained by pixel statistics. After calculation of the actual percentages of lesion areas in the areas of the corresponding whole diseased leaves based on the obtained pixel numbers, the training sets and testing sets were constructed for each severity class by using the system sampling method with two sampling ratios of 4:1 and 3:2. Then the mean and standard deviation of the actual percentages of lesion areas contained in each training set were calculated, respectively. For each sampling ratio, two methods, one based on the midpoint value of the means of the actual percentages of lesion areas corresponding to two adjacent severity classes and the other based on the distribution range of most of the actual percentages of lesion areas, were used to determine the midpoint-of-two-adjacent-means-based actual percentage reference range and the 90%, 95%, and 99% reference ranges of the actual percentages of lesion areas for each severity class. According to the determined reference ranges, the severity of each diseased leaf in the training sets and testing sets was assessed. The results showed that high assessment accuracies (not lower than 85%) for the training sets and testing sets were achieved, demonstrating that the proposed methods could be used to conduct severity assessment of wheat stripe rust based on the actual percentages of lesion areas. This study provides a reference for accurate severity assessments of plant diseases.
The timely and accurate identification of stripe rust and leaf rust is essential in effective disease control and the safe production of wheat worldwide. To investigate methods for identifying the two diseases on different wheat varieties based on image processing technology, single-leaf images of the diseases on different wheat varieties, acquired under field and laboratory environmental conditions, were processed. After image scaling, median filtering, morphological reconstruction, and lesion segmentation on the images, 140 color, texture, and shape features were extracted from the lesion images; then, feature selections were conducted using methods including ReliefF, 1R, correlation-based feature selection, and principal components analysis combined with support vector machine (SVM), back propagation neural network (BPNN), and random forest (RF), respectively. For the individual-variety disease identification SVM, BPNN, and RF models built with the optimal feature combinations, the identification accuracies of the training sets and the testing sets on the same individual varieties acquired under the same image acquisition conditions as the training sets used for modeling were 87.18–100.00%, but most of the identification accuracies of the testing sets for other individual varieties were low. For the multi-variety disease identification SVM, BPNN, and RF models built with the merged optimal feature combinations based on the multi-variety disease images acquired under field and laboratory environmental conditions, identification accuracies in the range of 82.05–100.00% were achieved on the training set, the corresponding multi-variety disease image testing set, and all the individual-variety disease image testing sets. The results indicated that the identification of images of stripe rust and leaf rust could be greatly affected by wheat varieties, but satisfactory identification performances could be achieved by building multi-variety disease identification models based on disease images from multiple varieties under different environments. This study provides an effective method for the accurate identification of stripe rust and leaf rust and could be a useful reference for the automatic identification of other plant diseases.
IntroductionThe accurate severity assessment of wheat stripe rust is the basis for the pathogen-host interaction phenotyping, disease prediction, and disease control measure making.MethodsTo realize the rapid and accurate severity assessment of the disease, the severity assessment methods of the disease were investigated based on machine learning in this study. Based on the actual percentages of the lesion areas in the areas of the corresponding whole single diseased wheat leaves of each severity class of the disease, obtained after the image segmentation operations on the acquired single diseased wheat leaf images and the pixel statistics operations on the segmented images by using image processing software, under two conditions of considering healthy single wheat leaves or not, the training and testing sets were constructed by using two modeling ratios of 4:1 and 3:2, respectively. Then, based on the training sets, two unsupervised learning methods including K-means clustering algorithm and spectral clustering and three supervised learning methods including support vector machine, random forest, and K-nearest neighbor were used to build severity assessment models of the disease, respectively.ResultsRegardless of whether the healthy wheat leaves were considered or not, when the modeling ratios were 4:1 and 3:2, satisfactory assessment performances on the training and testing sets can be achieved by using the optimal models based on unsupervised learning and those based on supervised learning. In particular, the assessment performances obtained by using the optimal random forest models were the best, with the accuracies, precisions, recalls, and F1 scores for all the severity classes of the training and testing sets equal to 100.00% and the overall accuracies of the training and testing sets equal to 100.00%.DiscussionThe simple, rapid, and easy-to-operate severity assessment methods based on machine learning were provided for wheat stripe rust in this study. This study provides a basis for the automatic severity assessment of wheat stripe rust based on image processing technology, and provides a reference for the severity assessments of other plant diseases.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.