【作者】 马铭；

【导师】 黄战华；

【作者基本信息】 天津大学 ， 光学工程， 2007， 硕士

【摘要】 随着现代交通的迅速发展,停车场资源越来越紧张。如何有效的管理停车场车位、提高停车场的使用效率,是智能交通系统面临的一个课题。与地感线圈和超声波探测技术相比,基于视频的车位监控方法具有节约硬件成本、易于安装施工、便于扩展其他附加功能,如停车存留登记、重要车辆看护等优点。针对某停车场内停车位多、车辆进出频繁的特点,本文设计出一套基于视频的停车场车位状态监控识别系统,通过对各停车位进行实时监控,得到每个车位的占用状态,实现停车场的高效管理。本文完成了以下几个方面的工作:1.根据停车场现场条件及车位分布要求,设计了视频采集光学系统,确定了光学系统的各参数,并根据车位状态监控的要求对光学系统进行了分析与评价;2.深入分析了视频监控的原理和光学模型,提出了车位监控的3个判据,结合实验得出了不同条件下的判据数值特征,确定了监控算法;3.以VC++6.0为开发平台,设计了一套基于视频的停车场车位状态监控识别软件,各模块分别实现了视频图像的获取、车位状态的监控、人机交互以及外部设备交互等功能;4.进行了基于AVI文件的监控实验,验证了监控算法和系统软件的可靠性,完成了软件和硬件的调试与改进。更多还原

【Abstract】 With the rapid development of modern traffic, the parking resource is facing great pressure. Among the projects of intelligent traffic system, it is crucial to effectively manage the park and accordingly improve its efficiency. Compared with the traditional methods upon induction coil or ultrasonic technology, the solution to parking spaces monitor based on computer vision has advantages in cost of hardware, convenience of setting. Moreover, it is apt to functions’evolution for parking registration or guarding the security of valuable vehicles.Since the great number of parking areas and frequent changes between entrance and exit in the park, the monitor system focuses on movement detecting and state recognition of each area, in order to improve efficiency of management.Works completed are as following:1. Based on the construction of park, scheme out and estimate the optics system for video capture according to the requirements of states monitoring.2. Analyze the principle of vision-based monitoring methods and the module of imaging, then import and refine 3 criterions as an algorithm, and finally distinguish varied characters among different situation.3. Finish the frame of software by MS VC++ 6.0, and realize the system which consists of video capture, automatic monitor, interface for user operation and communication with peripheral equipments.4. Experiment with AVI video files, test the reliability of algorithm and software, and complete debug works in both hardware and software.更多还原