Intelligent traffic monitoring using video data mining

点击翻译按钮获取中文摘要

Intelligent traffic monitoring systems need to provide continuous service with appropriate adaptation in various dynamic situations. To ensure performance guarantees, we need to have sufficient knowledge of the system and its properties. A transportation system needs to have sufficient knowledge of its situations and its subsequent changes, which can be obtained from video streams. It is necessary for the system to have a prior knowledge of its interactions with various objects of the environment involving different timing constraints to ensure safe and correct operations. It is also essential for the system to extract timing properties by mining video data, and understand the flow of execution through creating a knowledge base. Our aim is to create a knowledge base called VideoMine from the monitored video stream containing traffic information of various objects. Our objectives also include that accessing information from VideoMine will consume significantly less memory, allow faster information processing, and help to identify, analyze and validate safety aspects.