Hyperspectral image analysis based on BoSW model for rice panicle blast grading
详细信息 查看全文
- 作者：Shuangping Huanga ; b ; huangshuangping@gmail.com" class="auth_mail" title="E-mail the corresponding author ; Long Qib ; c ; qilong@scau.edu.cn" class="auth_mail" title="E-mail the corresponding author ; Xu Mab ; c ; maxu1959@scau.edu.cn" class="auth_mail" title="E-mail the corresponding author ; Kunnan Xueb ; xuekunnan@gmail.com" class="auth_mail" title="E-mail the corresponding author ; Wenjuan Wangd ; wangwj@gdppri.cn" class="auth_mail" title="E-mail the corresponding author ; Xiaoyuan Zhud ; zhuxy@gdppri.cn" class="auth_mail" title="E-mail the corresponding author
- 关键词：BoSW ; bag of spectrum words ; chi-SVM ; chi-square kernel support vector machine ; PCA ; principal component analysis ; ROI ; region of interest ; VIS ; visible ; NIR ; near infrared ; IRRI ; International Rice Research Institute
- 刊名：Computers and Electronics in Agriculture
- 出版年：2015
- 出版时间：October 2015
- 年：2015
- 卷：118
- 期：Complete
- 页码：167-178
- 全文大小：1907 K
文摘
- •
A new BoSW model for hyperspectral image data representation is proposed.
- •
High-dimension hyperspectral image is reduced to a compact vector by BoSW model.
- •
Spatial and full-spectral information are jointly considered.
- •
Spectrum prototype is searched and counted to reflect the severity of panicle blast.
- •
Classification accuracy is up to 81.41% for six and 96.40% for two infection levels.