摘要
针对卫星传感器成像的特点,提出了一种基于非下采样剪切波变换的脉冲耦合神经网络与稀疏表示相结合的卫星遥感图像融合方法。通过使用不同的融合规则对子带系数进行融合,得到信息更丰富的多光谱图像。为了验证该方法的有效性,实验过程使用了地球眼卫星数据和快鸟卫星数据,实验结果表明,该方法的实验结果无论从视觉效果还是客观评价指标上均优于对比方法得到的结果。
In the course of orbit,satellites are vulnerable to the influence of atmospheric resistance and solar pressure which results in the loss of image data obtained from satellites.According to the characteristics of satellite sensor imaging,a satellite remote sensing image fusion method based on the combination of pulse coupled neural network and sparse representation is proposed.The multispectral images with richer information are obtained by using different fusion rules to fuse subband coefficients.In order to verify the effectiveness of this method,the earth eye satellite data and fast bird satellite data are used in the experiments.The experimental results show that the experimental results of this method are better than those of the comparison method in terms of visual effect and objective evaluation indexes.
作者
曹义亲
杨世超
谢舒慧
Cao Yiqin;Yang Shichao;Xie Shuhui(College of Software,East China Jiaotong University,Nangchang 330013,China)
出处
《航天控制》
CSCD
北大核心
2020年第2期44-50,共7页
Aerospace Control
基金
国家自然科学基金项目(61663009)
江西省科技支撑计划重点项目(20161BBE50081)。
关键词
非下采样剪切波
卫星遥感图像融合
稀疏表示
脉冲耦合神经网络
Non-subsampled shearlet transform
Satellite remote sensing image fusion
Sparse representation
Pulse coupled neural network
