|
|||||
|
|
| 基于迷向圆变换和BP网络的不变识别方法 | |
| 引用本文： | 周升响,张兵,孟藏珍.基于迷向圆变换和BP网络的不变识别方法[J].空军雷达学院学报,2006,20(2):129-131,137. |
| 作者姓名： | 周升响 张兵 孟藏珍 |
| 作者单位： | 1. 空军雷达学院预警探测装备系,武汉,430019;空军雷达学院研究生管理大队,武汉,430019 2. 空军雷达学院预警探测装备系,武汉,430019 3. 空军雷达学院研究生管理大队,武汉,430019 |
| 摘 要： | 为提高对具有旋转、缩放和平移3种变形的二值图像的识别率和简化识别算法，提出利用迷向圆变换和图像面积调整方法将图像的3种变形归一化，再提取归一化图像的水平和垂直2个方向的统计特征输入BP神经网络进行分类识别，从而使神经网络的输入维数降低．仿真结果表明，该方法对具有上述3种变形的图像具有较高的识别率. |
| 关 键 词： | 迷向圆变换 面积调整 BP网络 |
| 文章编号： | CN42-1564（2006）02-0129-03 |
| 收稿时间： | 02 28 2006 12:00AM |
| 修稿时间： | 2006-02-282006-03-03 |
Invariance Recognition Method Based on Isotropic Circular Transform and BP Network |
|
| ZHOU Sheng-xiang,ZHANG Bing,MENG Cang-zhen.Invariance Recognition Method Based on Isotropic Circular Transform and BP Network[J].Journal of Air Force Radar Academy,2006,20(2):129-131,137. | |
| Authors: | ZHOU Sheng-xiang ZHANG Bing MENG Cang-zhen |
| Affiliation: | 1. Department of Early Warning Detection Equipment, AFRA, Wuhan 430019, China; 2. Department of Graduate Management, AFRA, Wuhan 430019, China |
| Abstract: | In order to improve the recognition rates for binary images with rotation, scale and translation deformations and to simplify the recognition algorithm, isotropic circular transform and area adjusting method were employed to normalize the rotation, scale and translation of the images, then the statistical features along the horizontal and vertical directions of the normalized images were abstracted and input to the BP neural network to be recognized, which reduced the input dimensions of the neural network. Simulation results show that the proposed method has higher recognition rates on images with the above-mentioned deformations. |
| Keywords: | |
| 本文献已被 维普 万方数据 等数据库收录！ | |