Current Issue Cover
自联想神经网络的遥感图象主分量提取

王耀南1,2, 谢 勇1,2, 毛建旭1,2, 李树涛1,2(1.湖南大学电气与信息工程系,长沙 410082;2.中国科学院模式识别国家重点实验室,北京 100080)

摘 要
提出了一种自联想神经网络的遥感图象主分量提取方法.这种方法可以应用于图象的压缩、特征提取和图象滤波中.实验结果表明:自联想神经网络算法简单、易于实现,其压缩效果与K-L变换相当.
关键词
Principal Component Extraction for Remote Sensing Image\nUsing Auto-Association Neural Network

()

Abstract
A new method of principal component extraction for remote sensing image based on auto-association\nneural network is presented in this paper. The proposed method is applied to the image data compression, feature\nextraction and filtering. Experiments show that auto-association neural network for extracting the principal\ncomponents of the image is as good as the conventional K-L transform.
Keywords

订阅号|日报