| 一种实用的边缘强度互相关匹配可信度分析方法 |
| |
| 引用本文： | 陈建祥,桑农.一种实用的边缘强度互相关匹配可信度分析方法[J].宇航学报,2005,26(6):722-725. |
| |
| 作者姓名： | 陈建祥 桑农 |
| |
| 作者单位： | 1. 北京西城区爱民街1号南楼电子信息基础部,100034
2. 华中科技大学图像识别与人工智能研究所,武汉,430074 |
| |
| 摘 要： | 匹配可信度是衡量匹配质量的重要指标，针对归一化边缘强度互相关匹配算法，研究了基于神经网络的匹配可信度判别方法。通过对参考图进行加噪实验获得训练样本，对BP网络进行训练，并将训练后的网络用于匹配可信度的判别。针对实际卫片与航片图像对的实验结果，平均判决正确率达到90％以上，证实了该方法的有效性。
|
| 关 键 词： | 归一化边缘强度互相关 匹配可信度 BP网 噪声模型 |
| 文章编号： | 1000-1328（2005）00-0722-04 |
| 收稿时间： | 04 15 2004 12:00AM |
| 修稿时间： | 2004-04-152005-07-04 |
A Practical Matching Confidence Analysis Technique Based on Neural Networks for Normalized Edge Magnitude Cross Correlation |
| |
| CHEN Jian-xiang,SANG Nong.A Practical Matching Confidence Analysis Technique Based on Neural Networks for Normalized Edge Magnitude Cross Correlation[J].Journal of Astronautics,2005,26(6):722-725. |
| |
| Authors: | CHEN Jian-xiang SANG Nong |
| |
| Institution: | 1. Base Department of Electronic Information, South Building, 1st Aiming Street, Westcity Beijing 100034, China; 2.Institute for Pattern Recognition and Artificial Intelligence, Huazhong University of Science and Technology, Wuhan 430074, China |
| |
| Abstract: | Matching Confidence is important for measuring the matching quality. For normalized edge magnitude cross correlation matching algorithm, the measure of matching confidence based on neural networks was studied. The training samples used to train a BP network were the matching results with the reference image and its noisy versions. The trained network can be adopted to measure matching confidence, Experimental results with real satellite images and aerial images gave the average correct rate over 90%, and proved the effectiveness of the method. |
| |
| Keywords: | Normalized edge magnitude cross correlation Matching confidence BP networks Noise models |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
