Скачать или смотреть Using Getis-Ord Gi* Analysis on Multispectral Drone Data

In this video, Jesse discusses slide deck provide by George Gaeke, who demonstrated the use of drones and multispectral imaging for vegetation assessments on cultivated sites at CLRA 2025. The focus of this video is to try look at Getis-Ord Gi* hotspot/coldspot analysis. The presenter experiments with Python code in Google Colab to perform this analysis on NDVI raster data, encountering challenges such as handling no data values and understanding different data types (float 32 vs. float 64). They emphasize the importance of understanding data well enough to validate AI tool outputs. After running the analysis, they identify cold and hot spots related to poor growth areas but express uncertainty about interpreting the results fully. The video concludes with thoughts on local versus global spatial autocorrelation and the potential benefits of identifying significant local areas amidst broader trends. Overall, it highlights an exploration of advanced analytical techniques in agricultural assessment using technology.