Satellites can now take videos, which contain more reflectance info than traditional images. Ye (Wuhan University) et al. examined the accuracy of classifying urban objects based on videos from the Jilin-1 agile video satellite constructed by Chang Guang Satellite Technology Co., Ltd. in Jinlin, China, using machine learning. Higher number of viewing angles in the training data led to greater classification accuracy when using a robust machine learning method like ensemble learning. The findings proved the value of satellite videos.

https://www.mdpi.com/2072-4292/14/10/2324/htm

#remote_sensing
#satellite_videos
#method_advancements
#pilot_scale