Nondestructive detection of sweet potato leaf curl virus using 3D hyperspectral imaging paired to machine learning

Authors: YANG, YICAN - Mississippi State University; WIJEWARDANE, NUWAN - Mississippi State University; SLONECKI, TYLER - Cornell University; Wadl, Phillip; Andreason, Sharon; CHEN, JINGDAO - Mississippi State University; HARVEY, LORIN - Mississippi State University

Submitted to: Smart Agricultural Technology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/10/2025
Publication Date: 5/11/2025
Citation: Yang, Y., Wijewardane, N.K., Slonecki, T.J., Wadl, P.A., Andreason, S.A., Chen, J., Harvey, L. 2025. Nondestructive detection of sweet potato leaf curl virus using 3D hyperspectral imaging paired to machine learning. Smart Agricultural Technology. https://doi.org/10.1016/j.atech.2025.101004.
DOI: https://doi.org/10.1016/j.atech.2025.101004

Interpretive Summary: Sweetpotato, is an economically important specialty crop in the United States that is susceptible to a range of diseases that can severely compromise yield and quality. Detecting these diseases, such as sweet potato leaf curl virus (SPLCV), is particularly crucial when storage roots are developing in the soil. SPLCV causes symptoms such as leaf curling, vein clearing, and chlorosis; however, these symptoms are often transient and disappear upon plant maturation. Visual inspection can be unreliable and potentially misleading without additional diagnostic tools. Accurate and early detection of SPLCV is crucial for effective management of this viral disease. The objective of our study was to investigate the potential of using 3D hyperspectral imaging to detect SPLCV. In our study, healthy and SPLCV-infected sweetpotato plants were imaged and machine learning models were used to distinguish healthy versus diseased plants. The results showed that the optimal model yielded a classification accuracy of 89.75%, exceeding the original data and other configurations. Overall, the use 3D hyperspectral imaging paired with machine learning for non-invasive diagnosis of SPLCV represents a promising step forward in agricultural technology, offering a viable solution for large-scale monitoring and management of crop health.

Technical Abstract: Sweet potato leaf curl virus (SPLCV) is a systemic viral disease of sweetpotato plants causing significant yield losses and posing a major threat to sweetpotato production. To achieve effective disease management, prompt detection of SPLCV is necessary and crucial. Currently, SPLCV is detected using a variety of molecular techniques, including polymerase chain reaction, loop-mediated isothermal amplification, and enzyme-linked immunosorbent assay, which are laborious, expensive, time-intensive, and limited in practicality for use in rapid, in situ diagnostic applications. Therefore, there is a need to establish alternative, point-of-care techniques for SPLCV detection. The goal of this study was to investigate the potential of using multispectral imaging and 3D-pointcloud data to detect SPLCV. In this study, 3D point cloud data derived from multispectral laser scanning of healthy and SPLCV-infected sweetpotato plants were modeled with the PointNet++ neural network to discriminate healthy versus diseased plants. Spatial coordinates (X, Y, Z) and color information (R, G, B) of sweetpotato leaf point cloud data were used to develop and improve the efficiency of the classification process. The collected 3D point cloud dataset was first randomly split as training and testing sets followed by downsampling to decrease the computational demand. Experiments were then conducted to fine-tune the hyperparameters of the PointNet++ algorithm. The results showed that the optimal hyperparameter configuration entailed the adoption of the multi-scale sampling and grouping (MSG) strategy, with the augmented point cloud down sampled to 2048 points and a batch size of 8, yielding a classification accuracy of 85.6%. The findings of our study showed the feasibility of using multispectral imaging and employing the deep learning model PointNet++ for non-destructive, rapid detection of SPLCV in sweetpotato.