Crop Water Stress Index Measurements and Mapping Using Image and Point Data for Upland Crop and Orchard Fruit Production

Reliable assessment of crop water stress is essential for improving irrigation efficiency for upland crop and orchard fruit production. Thermal sensing combined with the crop water stress index (CWSI) provides a non-destructive approach for characterizing plant water status; however, its performance depends strongly on baseline determination, sensing platform, canopy architecture, environmental conditions, image-processing procedures, and validation strategy. This paper aimed to review thermal imaging and point-based infrared thermometry (IRT) approaches for CWSI measurement and mapping, with particular emphasis on sensing requirements of upland field crops and orchard fruit tree canopies. A PRISMA-aligned literature search was conducted using major scientific databases and complementary citation tracking, resulting in 68 primary studies included in the structured thematic synthesis. Empirical, theoretical, and artificial-reference CWSI formulations were compared together with handheld, ground-based, unmanned aerial vehicle (UAV)-mounted, fixed camera, and point-based IRT sensing configurations. Radiometric calibration, canopy segmentation, temperature extraction, spatial interpolation, and physiological validation methods were critically evaluated. The reviewed evidence indicates that image-based approaches provide valuable spatial information but remain sensitive to radiometric drift, mixed canopy–background pixels, shadowing, and baseline uncertainty, whereas point-based IRT offers high temporal resolution but depends strongly on representative sensor placement and field-of-view geometry. These limitations differed according to canopy architecture, with soil–background contamination being particularly important in incomplete field canopies and radiation-driven within-canopy heterogeneity being more important in orchards. Combining periodic spatial thermal mapping with continuous point-based monitoring provided a promising framework for precision irrigation. Future research should prioritize transferable baseline parameterization, standardized uncertainty reporting, multi-season commercial validation, benchmark datasets, and crop-specific irrigation thresholds.