论文部分内容阅读
研究了可分解马尔科夫网的概念、方法,分析了可分解马尔科夫网在序列图像数据挖掘中的作用与意义,并直接将马尔科夫网的结构作为决策或推理依据,应用于问题求解,拓广了可分解马尔科夫网应用的可能性;以真实交通违章的视频图像为例,以多种粒度(节点数)广泛研究建立视频图像间的可分解马尔科夫网并分析其对问题的适用性,通过网络结构分析来检测视频图像间的差异,从而发现某种有意义的模式(例如,交通违章);仿真结果表明所提方法具有实用价值和较好效果;研究结果表明可分解马尔科夫网可以很好地揭示数据间的抽象近邻关系,并且这种网络具有很好的知识表达和逻辑推理的作用,是重要的模式识别方法。
The concept and method of decomposing Markov network are studied. The function and significance of decomposable Markov network in sequence image data mining are analyzed. The structure of Markov network is directly used as decision-making or inference basis and applied to the problem The possibility of decomposing Markov network is broadened. Taking the real traffic violation video as an example, the disassembly Markov network between video images is extensively studied with various granularities (nodes) and analyzed The applicability of the problem, through the network structure analysis to detect the difference between the video images to find some meaningful mode (for example, traffic violations); simulation results show that the proposed method has practical value and good results; the results show that Decomposable Markovian network can well reveal the abstract neighborhood between data, and this network has a good knowledge of the role of logical reasoning and is an important pattern recognition method.