摘要
本文运用神经网络BP算法对高分辨率航空多光谱图像进行了模式识别研究,成功地识别出注射过金属离子溶液树冠,并与最大似然法的分类结果进行了比较。得出在分类精度上,神经网络模式识别比传统的最大似然分类方法具有优势的结论。
This paper studied mode identification of aerial high resolution multispectral images and has successfully identified the crown injected metal ion solution. The comparison of classification results between neural network and maximum likelihood rule indicates that neural network is better in both of classification accuracy and rate.
出处
《遥感学报》
EI
CSCD
1998年第3期176-179,T001,共5页
NATIONAL REMOTE SENSING BULLETIN
基金
国家自然科学基金
关键词
神经网络
最大似然法
金属离子溶液
树冠
遥感
Neural network
Maximum likelihood rule
Mode identification
Metal ion
Crown