Potential and Limitations of Computer Vision for Crop Water Stress Detection in Irrigation Scheduling

Abstract

Computer Vision has become increasingly important in smart farming applications, including scheduling crop irrigation. A combination of various remote sensing devices enables continuous monitoring of a crop and non-destructive prediction of irrigation time. Appropriately scheduled and precisely targeted irrigation enables sustainable use of this limited resource. In agriculture, absorption-based and thermal-based imagery are used to monitor plant conditions through indices such as the Normalized Difference Water Index (NDWI) and Crop Water Stress Index (CWSI). This paper provides an overview of the concept and components of monitoring systems for automated irrigation scheduling. It explains the potential and limitations of applying computer vision-based systems for plant stress detection, providing insights to advance understanding in this growing field.

Code and data availability

This is an overview/review paper on computer vision for crop water stress detection. It describes the PlantSens prototype system and cites prior PlantSens publications, but no public phenotype dataset, image/sensor data, analysis code, trained model, or supplement with such assets is mentioned, and no data or code-avai

No evidence-backed public reproduction asset is currently recorded.