摘要
介绍了基于数据融合方法的高可靠性红外图像信号处理系统,分析了检测算法、阈值因子、检测概率和虚警率之间的关系,对比讨论了提高图像检测可靠性的两种方法,设计了基于算法多模型数据融合(Arithmetic -Multi-Model-Data-Fusion-Method,AMMDFM)的新方案。依据该方案所设计的系统,原理上保证了具有实时多目标处理能力,并且检测概率和虚警率指标大大优于以往常规硬件系统。可以在离线和在线两种方式下工作,方便了系统的调试,系统性能稳定,功耗低,体积小,集成度高,适用于安装在导弹等武器平台上作末制导寻的导引头。该方法是以增加系统硬件复杂程度为代价换取对红外图像目标检测可靠性的。
A more reliable infrared image information processing system based on data-fusion model is proposed . A new scheme based on AMMDF (Arithmetic-Multi-Model-Data-Fusion) Method has been designed, with the analysis of detecting arithmetic, threshold factor and detection probability to false-detecting probability and the discussion of two schemes for improving reliability of image detecting. The system designed according to AMMDFM has the capability of multi-aims processing in real-time, with better index in false-detecting and leaking-detecting probability than other general hardware systems. Debugging the system becomes easily, because it can work in two modes: on-line and off –line. With higher stability, lower power, smaller volume, greater integration, it can be used for terminal homing guidance of missiles or other weapons. This method trades increasing the complexity of hardware in the system for the detection reliability of infrared image targets .
出处
《红外技术》
CSCD
北大核心
2004年第6期17-21,共5页
Infrared Technology
基金
陕西省自然科学基金(合同号:2002F13)
西安电子科技大学青年基金(编号:03003)资助项目
