摘要
对空中目标物体实时跟踪目前主要采用雷达探测的方法,但时常会受到电磁场与电磁波的干扰,降低了探测的准确性,为解决此类问题,提出了一种基于空中视觉实时跟踪系统的解决方案。该方案适用于小范围空中动态物体的跟踪,包括动态目标的差值提取,二值图像的滤波和模板图像的实时更新,以及动态目标的特征提取、云台控制。为提高系统的准确检测性,提出了一种基于改进的动态模板匹配与帧间差分相结合的空中目标检测方法。该方法有效地解决了电磁场干扰、环境变化使动态目标无法捕捉的问题,根据实验表明,该系统能够很好地检测到空中动态物体,具有很好的鲁棒性、实时性和清晰度。
At present,the mainly method of real-time tracking to air targets is radar detection.But usually,this method will be affected by interference of electromagnetic field and electromagnetic.So the accuracy of the detection will be decrease.To solve such problems,a solution is put forward which based on air visual real-time tracking.The scheme is suitable for small range of air dynamic object tracking,including the extraction of dynamic target＇ s difference,the filtering of binary image,the real-time updates of template image,the feature extraction of dynamic target,and the cloud platform control.For increasing accuracy of the system＇ s detectability,an air target detection＇ s method is proposed what combine matching dynamic template which based on the improved one and difference between frames.This method will solve the electromagnetic interference effectively and the problem of dynamic target can＇ t be captured because of circumstances change.The experiment results show that the system can detect air dynamic object very well,and it possess good robustness,realtime performance and sharpness.
出处
《电视技术》
北大核心
2015年第11期67-70,共4页
Video Engineering
基金
广西研究生教育创新计划项目(2011105950804M30)
广西自然科学基金项目(2013GXNSFDA019002
2014GXNSFGA118003)
广西科学研究与技术开发项目(桂科合1346010-5)
桂林市科学研究与技术开发项目(20130122-1)
广西教育厅重点项目(ZD2014057)
关键词
空中视觉
云台控制
动态模板匹配
帧间差分
目标检测
air visual
cloud platform control
matching dynamic template
difference between frames
air target detection
