| 一种基于神经网络的超分辨图像重构方法 |
| |
| 引用本文： | 刘梅,刘慧念,刘伟东,王彦珍,许荣庆.一种基于神经网络的超分辨图像重构方法[J].哈尔滨建筑大学学报,2003(6). |
| |
| 作者姓名： | 刘梅 刘慧念 刘伟东 王彦珍 许荣庆 |
| |
| 作者单位： | 哈尔滨工业大学电子工程研究所，哈尔滨工业大学电子工程研究所，黑龙江省科技开发中心，北京大唐电信高鸿数据网络技术有限公司，哈尔滨工业大学电子工程研究所 黑龙江 哈尔滨 150001，黑龙江 哈尔滨 150001，黑龙江 哈尔滨 150001，北京 100088，黑龙江 哈尔滨 150001 |
| |
| 摘 要： | 针对在航空航天遥感领域中使用CCD相机对景物进行成像时,由干像元尺寸的限制导致图像分辨率低的问题,提出了采用人工神经网络映射图像重构过程非线性特性的方法。在相机参数已知的情况下,用后向误差传播(BP)神经网络融合从不同角度对同一景物重复拍照得到的多帧低分辨率图像的冗余信息,重构得到较高分辨率的图像。在模拟成像仿真实验中得到了分辨率提高4倍、信噪比接近30dB的超分辨结果。
|
| 关 键 词： | 图像重构 超分辨 神经网络 共轭梯度 |
Reconstruction of super-resolution image based on neural network |
| |
| LIU Mei,LIU Hui-nian,LIU Wei-dong,WANG Yan-zhen,XU Rong-qing,. Research Institute of Electronic Engineering Techniques,Harbin Institute of Technology,Harbin ,China
. Heilongjiang Science and Technology Development Center,Harbin ,China.Reconstruction of super-resolution image based on neural network[J].Journal of Harbin University of Civil Engineering and Architecture,2003(6). |
| |
| Authors: | LIU Mei LIU Hui-nian LIU Wei-dong WANG Yan-zhen XU Rong-qing Research Institute of Electronic Engineering Techniques Harbin Institute of Technology Harbin China
Heilongjiang Science and Technology Development Center Harbin China |
| |
| Affiliation: | LIU Mei1,LIU Hui-nian1,LIU Wei-dong2,WANG Yan-zhen3,XU Rong-qing1,1. Research Institute of Electronic Engineering Techniques,Harbin Institute of Technology,Harbin 150001,China
2. Heilongjiang Science and Technology Development Center,Harbin 150001,China
3. Beijing Datang Gaohong Network Company,Beijing,100088,China |
| |
| Abstract: | The reconstruction of super-resolution image based on Back - Propagation Neural Network algorithm is proposed to provide a higher quality reconstructed image with resolution improved by 4 times to avoid error accumulating of CG iterations, and get high-resolution image by using Back-Propagation Neural Network to map the nonlinear processing of reconstruction. It differs from those methods based on equation iterative solu- tion and can greatly improve the quality of super-resolved image. |
| |
| Keywords: | image reconstruction super-resolution neural networks conjugate gradient |
| 本文献已被 CNKI 等数据库收录！ |
|
