【作者】 徐鹏；

【导师】 韩少军；

【作者基本信息】 武汉理工大学 ， 机械电子工程， 2008， 硕士

【摘要】 机构的运动状况对于分析整个机械系统的运转有重要意义。对于动态对象,难以寻找固定的参考点,各种传统的位移监测方法都有其局限性,因此发展一种远距离、非接触的观测手段,并且能实现自动记录、处理连续数据的实时系统是非常必要的工作。在计算机技术的高速发展的基础上,图像处理技术也得到了快速的进步,计算机的高速运算能力为数字图像处理提供了技术保证。借助计算机,应用数字图象处理的方法对平面机构的运动状态进行实时检测,测量结果就可以直接的反映这个机械系统的工作状况。通过对摄像机拍录的图象序列进行自动分析来对被监控场景中的变化进行定位、识别和跟踪,并在此基础上分析和判断有关目标的行为。但实际的场景往往是错综复杂、变化无常的,而准确地对运动目标检测和跟踪是图象处理法能否用于机构运动分析的关键。本文利用视觉信息在时间和空间上的相关性,主要研究运动目标检测和跟踪的相关模型及算法。本文的另一项主要工作是对从序列图象中检测出的运动目标进行轨迹跟踪。通过建立目标跟踪的卡尔曼滤波器模型,对运动目标进行运动估计,从而减少噪声干扰和特征提取的搜索范围,提高了算法的效率。在目标匹配时,采用质心和面积相结合的方法,兼顾了目标的位置和形状,并利用匹配函数来进行目标匹配,取得了良好的效果。另外,算法中检测结果和跟踪结果的相互校正进一步提高了系统的稳定性。实验结果表明,目标跟踪的卡尔曼滤波模型比较符合实际的情况,跟踪算法能可靠地预测与跟踪出目标的运动轨迹。与一般物体速度测量方法相比,基于数字图象处理的平面连杆机构运动分析具有实时性,是一种远距离的、非接触式的测量方法,可用于比较恶劣的工作环境,所以有广泛的发展空间。更多还原

【Abstract】 Mechanism movement status has the vital significance to analysis entire mechanical system revolution. All kinds of traditional methods about displacement detection to dynamic object have their limitation, it is hard to find fixed reference point. So it is very necessary to develop a kind long distance and untouched detection system which can record date by itself and deal with a series of date in time. On the base of computer technology high speed development , The image processing technology also obtain great progress , The computer high speed operational capability provide the technical guarantee for digital image process. By the aid of computer, detective the plane mechanism movement status with digital image processing method. The measurement result may the directly reflect this mechanical system working condition.Through out locate, identify and track the image sequence which obtained by digital camera, and then understanding the behavior of moving objects. The actual scenes are usually complex and changeable. So detecting and tracking of moving objects is the key to analysis mechanism motivation. Base on the visual information of the time and space relationships, moving objects detection and tracking methods are studied in this thesis.Another major work of this thesis is to track moving objeets’ trajectory in image sequences obtained from the proposed detection algorithm .An objects tracking model based on Kalman filters is presented .The motion estimation based on Kalman filters can reduce noise disturbance and shrink the search range of characteristic and improved tracking efficiency. In the proeess of objects matehing, location and shape of objects are considered together, a matching function integrates the centroid and area of object is introduced to match objects between two consecutive image sequences. In addition, the emendation between the detection results and the tracking results improves stability of the whole tracking system better. Results of experiments show that the proposed tracking model based on Kalman filter relatively conforms to practical conditions and the tracking algorithm can reliably forceast and track the motion trajectories of objects. Compared with the general object velocity measurement method, the plane link mechanism movement analysis based on digital image processing has real-time character, is one kind of long-distance , and non-contact measuring technique, also can available in quite bad working conditions, so it can be obtained widespread development space.更多还原