@inproceedings{54c63e3692104b4da962b60f92856a9f,
title = "Identification of Phragmites australis and Spartina alterniflora in the Yangtze Estuary between Bayes and BP neural network using hyper-spectral data",
abstract = "The aim of this work was to identify the coastal wetland plants between Bayes and BP neural network using hyperspectral data in order to optimize the classification method. For this purpose, we chose two dominant plants (invasive S. alterniflora and native P. australis) in the Yangtze Estuary, the leaf spectral reflectance of P. australis and S. alterniflora were measured by ASD field spectral machine. We tested the Bayes method and BP neural network for the identification of these two species. Results showed that three different bands (i.e., 555 nm{\"i}1/4711 nm and 920 nm) could be identified as the sensitive bands for the input parameters for the two methods. Bayes method and BP neural network prediction model both performed well (Bayes prediction for 88.57\% accuracy, BP neural network model prediction for about 80\% accuracy), but Bayes theorem method could give higher accuracy and stability.",
keywords = "BP neural network, Bayes, Classify, Hyper-spectral data, Reflectance",
author = "Pudong Liu and Jiayuan Zhou and Runhe Shi and Chao Zhang and Chaoshun Liu and Zhibin Sun and Wei Gao",
note = "Publisher Copyright: {\textcopyright} COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.; Remote Sensing and Modeling of Ecosystems for Sustainability XIII ; Conference date: 31-08-2016",
year = "2016",
doi = "10.1117/12.2236772",
language = "英语",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "SPIE",
editor = "Wei Gao and Ni-Bin Chang",
booktitle = "Remote Sensing and Modeling of Ecosystems for Sustainability XIII",
address = "美国",
}