Early Detection of Stripe Rust in Winter Wheat Using Deep Residual Neural Networks

Abstract

Stripe rust (Pst) is a major disease of wheat crops leading untreated to severe yield losses. The use of fungicides is often essential to control Pst when sudden outbreaks are imminent. Sensors capable of detecting Pst in wheat crops could optimize the use of fungicides and improve disease monitoring in high-throughput field phenotyping. Now, deep learning provides new tools for image recognition and may pave the way for new camera based sensors that can identify symptoms in early stages of a disease outbreak within the field. The aim of this study was to teach an image classifier to detect Pst symptoms in winter wheat canopies based on a deep residual neural network (ResNet). For this purpose, a large annotation database was created from images taken by a standard RGB camera that was mounted on a platform at a height of 2 m. Images were acquired while the platform was moved over a randomized field experiment with Pst-inoculated and Pst-free plots of winter wheat. The image classifier was trained with 224 × 224 px patches tiled from the original, unprocessed camera images. The image classifier was tested on different stages of the disease outbreak. At patch level the image classifier reached a total accuracy of 90%. To test the image classifier on image level, the image classifier was evaluated with a sliding window using a large striding length of 224 px allowing for fast test performance. At image level, the image classifier reached a total accuracy of 77%. Even in a stage with very low disease spreading (0.5%) at the very beginning of the Pst outbreak, a detection accuracy of 57% was obtained. Still in the initial phase of the Pst outbreak with 2 to 4% of Pst disease spreading, detection accuracy with 76% could be attained. With further optimizations, the image classifier could be implemented in embedded systems and deployed on drones, vehicles or scanning systems for fast mapping of Pst outbreaks.

Combined UAV‐ and tractor‐based stripe rust monitoring in winter wheat under field conditions

“…In the classified images, there were still some misclassified pixels. Other image analyzing methods, especially machine‐learning approaches (e.g., Schirrmann et al., 2021), may lead to a lower number of those pixels. A requirement for using this method is that a prior comprehensive annotation of the training images is necessary.…”

A Deep-Learning-Based Approach for Wheat Yellow Rust Disease Recognition from Unmanned Aerial Vehicle Images

“…The BPNN algorithm had a simpler structure, and most of the misclassification results were in the vicinity of different classes of image elements, while the PSPNet algorithm had a cleaner classification result with the highest overall accuracy of 98%. Schirrmann, Michael et al [ 40 ] used the ResNet model to identify wheat stripe rust images taken by RGB cameras with a total accuracy of 90%. Hayit, Tolga et al [ 41 ] artificially planted wheat and inoculated with Urediniospore suspension, then used a camera to photograph wheat leaves to create a dataset, and identified them based on the Xception model.…”

Proof of concept study: a novel mobile in-canopy imaging system for detecting symptoms of fungal diseases in cereals

“…In addition, a combination of UAVs and groundbased imaging for species-specific disease symptom recognition has been conducted (Bohnenkamp et al 2019;Dammer et al 2021). Sophisticated image analysis methods, e.g., using machine-learning algorithms (Behmann et al 2015;Schirrmann et al 2021), are applied progressively to identify symptoms within the images automatically. Regarding to economic and ecological aspects, for a demand related precise crop protection a spraying of uninfected areas makes no sense.…”