Maintaining Optimum Closeup in Wheat FHB Detection Using 360-Degree Deep Scanning Method

摘要

<b><sc>Abstract.</sc></b> Fusarium Head Blight (FHB) results in massive yield and quality losses in wheat and barley annually. Evaluating and estimating the infection level on FHB-resistant lines is time-consuming, labor-intensive and needs expertise. This paper describes an innovative method to detect and assess the stage of FHB disease in wheat and barley by implementing the advancements in Artificial Intelligence (AI) and image processing and utilizing an innovative system. The application of image processing in precision agriculture due to the high-quality imagery techniques coupled with modern algorithms and the increased feasibility of fusing satellite imagery with information from sensors positioned in fields is promising and growing rapidly. Utilizing such technologies for accurate and early FHB disease detection can help substantially improve the wheat and barley breeding programs‘ efficiency. An innovative “360° phenotyping robot” has been designed by the Precision Ag team and fabricated in SDSU. In this study, the phenotyping robot was used to shoot images at a prespecified perspective to get an optimum closeup for identifying the FHB disease even at an early stage with minimum symptoms. The wheat images from an FHB-inoculated wheat field at the SDSU research farm in Volga, SD, were captured to train the proposed model. In addition to healthy and unhealthy plants, these images encompass various stages of the disease. The captured images were used to create the Convolutional Neural Network (CNN) data set to train a model to accurately detect the FHB in the early stages.