Fig. 6: Detecting wildebeest across different landscapes with variation in wildebeest spatial clustering patterns. | Nature Communications

Fig. 6: Detecting wildebeest across different landscapes with variation in wildebeest spatial clustering patterns.

From: Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape

Fig. 6: Detecting wildebeest across different landscapes with variation in wildebeest spatial clustering patterns.The alternative text for this image may have been generated using AI.

The figures in the first column show the detected wildebeest (red circles). The second column is a zoom of the imagery covered by the white square in the first column. a Detected wildebeest in GeoEye-1 imagery acquired on August 11th, 2009. In the zoomed-in image, the wildebeest are crossing the road near a dry riverbed. b Detected wildebeest in GeoEye-1 imagery acquired on August 10th, 2013. Wildebeest herd in open grasslands. c Detected wildebeest in WorldView-3 imagery acquired on July 17th, 2015. The wildebeest prepare to cross the Mara River. d Detected wildebeest in GeoEye-1 imagery acquired on August 2, 2018. Herds of wildebeest avoid the closed woodlands. e Detected wildebeest in WorldView-2 imagery acquired on October 8th, 2020. The wildebeest herds move through open woodlands and grasslands. These examples also show the heterogeneity between the satellite images, inclusive of spectral variation and different levels of contrast between the wildebeest and the background. Satellite image © 2009–2020 Maxar Technologies.

Back to article page