| 基于简约HJ神经网络的图像满幅校正 |
| |
| 引用本文： | 林家骏,杨明福.基于简约HJ神经网络的图像满幅校正[J].计算机应用与软件,2001,18(5):49-53. |
| |
| 作者姓名： | 林家骏 杨明福 |
| |
| 作者单位： | 华东理工大学信息技术与工程学院, |
| |
| 摘 要： | 文中提出一种简约HJ神经网络（RHJNN），用于对旋转变形的图形进行校正。该神经网络具有较简单的结构，数值计算与理论分析表明该网络更具鲁棒性。相关的满幅校正策略是实现无差后续图像处理的必要与有效手段。
|
| 关 键 词： | 盲分离 神经网络 鲁棒性 图像满幅校正 图像处理 图像识别 |
FULL SHAPE RESTORATION BASED ON REDUCED HJ NEURAL NETWORK |
| |
| Lin Jiajun Yang Mingfu.FULL SHAPE RESTORATION BASED ON REDUCED HJ NEURAL NETWORK[J].Computer Applications and Software,2001,18(5):49-53. |
| |
| Authors: | Lin Jiajun Yang Mingfu |
| |
| Abstract: | In this paper, the Reduced HJ Neural Network(RHJNN) is proposed to realize restoration for a rotated graph. The structure of RHJNN is simpler and its robustness is proved by numerical calculation and theoretical analysis. The related strategy of full shape is a necessary and efficient method to realize restoration without error. |
| |
| Keywords: | Restoration Blind separation Neural network Robust |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
