Hyperspectral imaging Vis-NIR and SWIR fusion for improved drought-stress identification of strawberry plants

Authors: FAQEERZADA, MOHAMMAD - Chungnam National University; KIM, HANGI - Chungnam National University; Kim, Moon; Baek, Insuck; Chan, Diane; CHO, BYOUNG-KWAN - Chungnam National University

Submitted to: Computers and Electronics in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/22/2025
Publication Date: 6/26/2025
Citation: Faqeerzada, M.A., Kim, H., Kim, M.S., Baek, I., Chan, D.E., Cho, B. 2025. Hyperspectral imaging Vis-NIR and SWIR fusion for improved drought-stress identification of strawberry plants. Computers and Electronics in Agriculture. 237 Part C. Article e10702. https://doi.org/10.1016/j.compag.2025.110702.
DOI: https://doi.org/10.1016/j.compag.2025.110702

Interpretive Summary: The ability to detect plant stresses before visible symptoms are apparent can be critical to effective management to prevent unrecoverable crop losses. Although a variety of hyperspectral methods have been investigated for the detection of plant stresses such as disease or drought, most such work has focused on the use of single-sensor spectral or imaging systems for detection, such as visible/near-infrared (VIS-NIR) or short-wave infrared (SWIR) imaging. This study investigates the use of hyperspectral fusion to improve the accuracy of spectral-imaging based detection of drought-stressed strawberry plants. Control (unstressed), drought-stressed, and recovering drought-stressed plants were imaged with two different hyperspectral line-scan imaging systems, one VIS-NIR and one SWIR, and moisture content measurements were acquired for leaves and soil during the periods of stress and recovery. A variety of spectral image processing methods were were performed in order to develop and compare classification models based on partial least-squares discriminant analysis that utilized as spectral inputs only VIS-NIR, only SWIR, or fused VIS-NIR and SWIR hyperspectral image data. The results showed that a classification model using fused spectral data with less data preprocessing could outperform models that used only VIS-NIR or only SWIR images combined with a variety of pre-processing methods commonly used to improve spectral-based classifications. The fused model demonstrated over 99% accuracy in classifying control, stressed, and recovery stage strawberry plants, illustrating the potential of hyperspectral image fusion for highly effective detection of drought stress in strawberry plants that could help fruit production managers prevent or mitigate strawberry crop losses due to drought stress.

Technical Abstract: Hyperspectral imaging systems that operate in the visible-near infrared (VIS-NIR) and short-wave infrared (SWIR) spectral regionsspectrums are increasingly recognized as practical and effective tools for enhancing crop management. HoweverDespite , their utility, hyperspectral systems can have some have limitations when focusing on specific spectral ranges, particularly their for spatial and spectral resolution. Image fusion techniques combine combining information from different sensors to enhance hyperspectral data can significantly improve, resulting in improved spatial and spectral resolution. This study presents advancements in hyperspectral image fusion achieved by using two line-scan sensors, one for VIS-NIRin the VIS-NIR range (397-1003 nm) and the other for SWIRin the SWIR range (894-2504 nm), to detect asymptomatic drought stress in strawberry plants. The images from both hyperspectral imaging systems were aligned based on feature and intensity combined with various geometric transformations for fusion. The resulting fused hyperspectral cube contained 403 bands covering a broad spectrum from 397 to 2500 nmcovers a broad spectrum from 397 to 2500 nm, containing 403 bands. Given the vulnerability of strawberry plants to drought, which can significantly affect both significantly affects their growth and yield, this study aimed to explore the potential of hyperspectral image fusion for high-throughput detection of drought-stressed strawberry plants. The fused images improved the performance of the PLS-DA detection model, increasing classification accuracy by up to 10%, achieving 99% accuracy in the prediction set, and reducing error rates compared to independently generated models. Spectral image fusion of image data from the VIS-NIR and SWIR regions two different wave ranges provides complementary information on plant physiology, biochemistry, and morphology before visible symptoms of stress appear.