摘要
提出一种基于协同理论的交通状态自动识别的新方法。首先,针对交通图像受光线、抖动干扰问题,提出了实时信息融合的动态原型选择模型;然后,利用直接求相似度重构序参量,实现了实时动态原型的选择;最后,依据快速协同网络框架,避免动力演化过程迭代,快速识别。协同方法可实现不同时段各种交通状态的自动检测。试验结果可说明该方法效果理想、速度快,具有较强的鲁棒性。
To put forward a new method of automatic recognition based on synergetics in traffic state. Firstly, according to crossing traffic image, set up the dynamic prototype selection model of real time information syncrotics. Secondly, using resemblance of parameter, to carry out selection of dynamic prototype. Lastly, to carry out an automatic recognition algorithm. Automatic detection of traffic states was achieved. The results showed that an effective synergetic recognition method has been developed.
出处
《中国体视学与图像分析》
2007年第1期37-42,共6页
Chinese Journal of Stereology and Image Analysis
基金
国家自然科学基金项目(60675058)
福建省自然科学基金项目(A0510010)
关键词
协同方法
交通状态
自动识别
synergetic method
traffic state
automatic recognition
