作者: Henning Gerstmann , Cornelia Gläßer , Detlef Thürkow , Markus Möller
DOI: 10.1007/S41064-018-0043-6
关键词:
摘要: Agricultural monitoring and assessment based on satellite data increasingly gains importance due to the growing number of available sensors with high geometric temporal resolution. Such tasks often require multiple images acquired specific dates that among others account for inter-annual phenological variations provide accurate results. This contribution presents an approach links peaks spectral separability profiles crop phases. The phases are spatially interpolated using a model ground observations. show respective development F-measure which is used as indicator class-wise separability. It originates from binary classifications vegetation indices computed each set archive covering years. Acquisition dates, repeatedly maximum define Potential alternative can be also defined. Experiments multi-temporal RapidEye imagery were performed three crops at two German test sites under different environmental conditions. results showed yellow ripeness, heading flowering function winter barley, wheat rapeseed. We could identify least identical, stable per type both sites, suggests transferability robustness presented approach.