摘要
Automatic recognition system for coccoliths (Système de Reconnaissance Automatique de Coccolithes, SYRACO) is a tool for automatic identification and count of coccolith by artificial neural networks. This software is able to identify individual coccoliths reliably at the same time as it is able to recognize almost all coccoliths present in a field of view, which would improve work efficiency significantly and prevent manual errors in routine work. The trained SYRACO system could identify 14 Quaternary coccolith taxa. The authors employ SYRACO to analyze coccolith specimens at Site MD05-2901 in the South China Sea and obtain the percentages of Florisphaera profunda, Emiliania huxleyi and Gephyrocapsa oceanica. The incorporation of the automatic recognition results and previous manual counts proves the utility of SYRACO on paleoceanographic research.