Crop classification with landsat multispectral scanner data II

摘要

The Landsat multispectral scanner data have been analyzed with a view toward crop identification and inventory. The notion of spectral-temporal trajectory as crop signature, introduced in an earlier paper, has been further developed leading to a formulation of the problem similar to one of recognition of hand-written characters. A simple classification rule based on angular features of the trajectory is discussed and classification results are given for Landsat images from several sites.