Shadow Cast Tracking for Deduction of Elevation Data Through Affine Matching Methods on Optical Satellite Imagery

Cast shadow from mountain ridges can be tracked over time. Through simple trigonometry this can give a clue about relative elevation change. These shadow ridges can be easily identified in multi -spectral satellite imagery as they are strong illumination changes in all spectral bands. Following shadow cast can be more complicated, due to different sun angles and changes in terrain slope. This results in a deformed appearance, which complicate tracking when simple pattern matching techniques are used. In this contribution we explore two affine matching methods in both the spatial and frequency domain, to increase the matching success of shadow cast on mountain glaciers.