基于人工神经网络和经验知识的遥感信息分类综合方法

REMOTE SENSING INFORMATION CLASSIFICATION BASED ON ARTIFICIAL NEURAL NETWORK AND KNOWLEDGE

提出基于三维 Hopfield人工神经网络模型的遥感信息分类综合算法将遥感图象分类与分类结果平滑过程有机地结合在一起 ,并方便地引入了经验知识指导网络进化 .实验表明该方法可明显提高森林类型划分。

This paper presents a remote sensing information classification method based on 3 D Hopfield neural network.It combines the classification process with the smoothing process ideally,and guides the evolution of network with knowledge.The results of experiments indicate that the presented method can enhance the classification precision remarkably.