| Summary: | Unmanned Aerial Vehicles (UAVs) equipped with thermal and multispectral imaging capabilities, which offer high spatial and temporal resolutions, are becoming increasingly valuable for timely crop monitoring and informed decision-making in precision agriculture. However, processing and extracting useful information from UAV images is often complex, time-consuming, and requires specialized software, which limits its broader adoption for practical field implementations. To address these challenges, the Agriculture Sensing and Artificial Intelligence (AgriSenAI), a user-friendly Python-based desktop application, was developed to automate processing and information extraction from UAV-acquired thermal and multispectral imagery. AgriSenAI was developed by integrating advanced image processing with geospatial analysis to streamline field and plot extraction, plant canopy detection, noise removal, and extraction of information at pixel, plot, and field scales. The application was designed and tested using UAV-based thermal and multispectral imagery collected daily for three years from a research field at the University of Florida's Tropical Research and Education Center in Homestead, Florida. The research field consisted of 12 plots of green beans and 12 plots of sweet corn. The processing time and accuracy of AgriSenAI were evaluated. Results showed that AgriSenAI had a very high level of accuracy in extracting pixel values and significantly reduced processing time and costs compared with traditional approaches involving commercial software. The streamlined AgriSenAI workflow produced reliable canopy temperature information and vegetation indices, demonstrating the capacity to handle large-scale datasets and enhance precision agriculture through improved efficiency and accuracy in remote sensing data processing and information extraction, which could potentially be used to inform timely and data-driven crop management decisions.
|
|---|
