摘要
针对当前嵌入式智能视频分析系统图像易抖动、漏报和误报率高、检测规则少等现状,提出了通过在视频服务器端应用视频稳定模块、采用多高斯分布的算法和引入规则引擎等方法的嵌入式智能视频分析系统设计方案。并通过交通流量检测实践结果验证该设计方案提高了视频图像防抖能力,降低了漏报和误报率,方便用户实现检测规则的灵活设置。
In view of shortage of current embedded intelligent video image analysis system, such as image dithering, high rate of false alarm and missing alarm, little rules setting and so on, this paper designs a solution of embedded intelligent video analysis system, through application of video stability module, adoption of mixture of Gaussians and usage of rules engine in video server. Practice verifies this solution can improve capability of protection image dithering, reduce rate of false alarm and missing alarm and set rules flexibly for user through a traffic flow detection case.
出处
《电子设计工程》
2016年第22期36-38,46,共4页
Electronic Design Engineering
基金
国家自然科学基金项目(61271333)
关键词
嵌入式系统
智能视频分析
视频监控
embedded system
intelligent video analysis
video surveillance
