论文部分内容阅读
遥感影像分类专家系统是遥感分类研究中的一个重要发展方向，然而，传统的统计模式识别法和人工神经网络分类法除了能完成具体的影像分类外，不能提供易于被人类理解的分类知识。文中介绍一种基于扩张矩阵的示例学习方法，并将其应用于遥感影像分类知识的自动获取。
Remote sensing image classification expert system is an important development direction in the research of remote sensing classification. However, traditional statistical pattern recognition and artificial neural network classification can not provide classification knowledge which is easy to be understood by human except for the specific image classification. In this paper, an example learning method based on expansion matrix is introduced and applied to the automatic acquisition of remote sensing image classification knowledge.