节点文献

基于超声波与机器视觉的自动泊车系统设计

Automatic Parking System Design Based on Ultrasonic and Machine Vision

【作者】 王文飞

【导师】 周泓;

【作者基本信息】 浙江大学 , 电子信息技术及仪器, 2011, 硕士

【摘要】 自动泊车系统是目前智能汽车领域的一个研究热点与难点,能够实现狭窄泊车位下的自动泊车,大大降低泊车的负担。本文首先对实现自动泊车系统涉及到的两个核心算法做了探索与研究,详细阐述了这两个算法的原理。提出参考车泊车位模型和泊车线车位模型,分别通过超声波测距与视频分析的方法实现泊车位搜索。根据车身运动轨迹的几何关系将泊车路径分成4个阶段,并计算出每个阶段的运动参数。接着提出自动泊车系统的总体方案,该包含传感信号单元、车控单元、信号处理单元。最后在电动模型车上构建了一个完整的嵌入式系统进行算法与方案的验证。该系统包含了传感信号电路、车控电路、信号处理电路这三大部份,对于每一个部分都详细介绍了该系统的硬件方案、软件构架。通过对所设计的嵌入式系统的测试与评估,验证了本文提出的算法和解决方案都是行之有效的。

【Abstract】 The automatic parking system is a hot spot in the smart car field. It can park the car automatic within a narrow space. That reduces the burden of parking car for driver. First, this theis researches two core algorithms for automatic parking system, and explain the principles of the algorithms in detail. Reference model car parking spaces and parking spaces line model, which is searched through the ultrasonic and the video analysis respectively, is presented. According to the geometric relationship of the path of the park, the parking procedure is divided into four phases. The parameters of each phase will be calculated.Second, the solution of automatic parking system is presented. The solution contains three units, including sensor signal unit, vehicle control unit, signal processing unit. Finally, a real embedded system, to verify the algorithms and solution, is completed based on an electric model car. The system includes a sensor signal circuit, the car control circuit, the signal processing circuit. For every part of that, detailed hardware and software architecture is explained.Through the test and evaluation for the embedded system, the proposed algorithms and solutions are proved effective.

  • 【网络出版投稿人】 浙江大学
  • 【网络出版年期】2011年 07期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络