Semi-Automated Analysis of Digital Photographs for Monitoring East Antarctic Vegetation

摘要: Climate change is affecting Antarctica and minimally destructive long-term monitoring of its unique ecosystems vital to detect biodiversity trends, understand how these communities. The use automated or semi-automated methods especially valuable in harsh polar environments, as access limited conditions extreme. We assessed moss health cover at six time points between 2003 2014 two East Antarctic sites. Semi-automatic object-based image analysis (OBIA) was used classify digital photographs using a set rules based on red, green, blue (RGB) hue-saturation-intensity (HSI) value thresholds, assigning vegetation categories healthy, stressed moribund lichens. Comparison with traditional visual estimates showed that percent OBIA classification fell within the range variation determined by methods. Overall health, mean percentages mosses quadrats, changed over 11 years both A marked increase stress decline observed across sites 2008, followed recovery baseline levels one site, but significantly more remaining communities other site. Our results confirm can be reliably estimated OBIA, providing similar accuracy estimation experts. resulting provide sensitive measure assess have informed conceptual framework for changing condition mosses. In demonstrating this method monitor ground small scales, we suggest it may also suitable extreme environments where repeat via images required.