High-resolution satellite imagery offers promising tools for wildlife monitoring. We evaluated the feasibility of detecting three large wintering waterbird species: White-naped Crane (Grus vipio), Whooper Swan (Cygnus cygnus), and Eurasian Spoonbill (Platalea leucorodia) using pansharpened imagery with spatial resolutions of 0.7 m and 0.55 m from KOMPSAT-3 and KOMPSAT-3A satellites. Morphometric analysis showed that G. vipio and C. cygnus averaged 66.4 cm and 65.3 cm in body length, corresponding to 1 to 2 pixels in the imagery. Bright objects were consistently detected near agricultural fields and reservoir margins, and C. cygnus was clearly identified based on spatial overlap with field observations. In contrast, G. vipio was not detected, likely due to low visual contrast between its gray plumage and the background, underscoring the role of chromatic contrast in image-based species detection. Bale silage, a potential source of confusion, was distinguishable based on shape and fixed location. The spatial distribution of detected objects aligned closely with Google Earth Pro imagery and field surveys, although the total number of detected individuals (n = 245) differed from ground counts. This discrepancy likely reflects resolution and background complexity. Improving detection accuracy may increase the utility of satellite imagery for habitat monitoring and conservation planning.



Keywords : High-resolution remote sensing, object detection, Avian distribution monitoring