基于BP神经网络的多源桥梁毁伤信息特征层融合  

The Feature-Level Fusion of the Multi-Image Damage Information about the Bridge Based on the BP Neural Network

作　　者：朱亚红[1] 汪民乐[1] 敬斌[1] 

机构地区：[1]第二炮兵工程学院,陕西西安710025

出　　处：《战术导弹技术》2012年第2期116-121,共6页Tactical Missile Technology

摘　　要：基于BP神经网络,建立了多源毁伤信息特征层融合系统,对多源桥梁毁伤信息进行BP网络特征层融合,首先对多源图像信息采用轮廓分析的目标毁伤特征提取算法,进行边缘及直线特征提取,其次进行毁伤区域不变矩特征关联,最后利用BP神经网络进行训练并测试,最终融合得到准确而清晰的桥梁毁伤信息,能够为导弹打击效果的毁伤评估提供准确的信息支撑。The feature-level fusion system about the multi-damage information is established with the BP neural network.The feature-level fusion of the multi-image damage information about the bridge is finished with the BP neural network.Firstly,a feature extracting computational method of the target damage on the profile analysis is proposed;secondly,the feature associating method based on the invariant moment in the image zone is taken.The accurate information which can provide the accurate information for the damage assessment of the missile striking effectiveness is obtained.

关 键 词：BP神经网络 特征层融合 桥梁毁伤 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]