基于DSP的复杂背景下运动目标追踪系统的研究

【作者】 李海明；

【导师】 陈新；

【作者基本信息】 福州大学 ， 通信与信息系统， 2002， 硕士

【摘要】 近年来，在复杂背景下识别和跟踪运动目标、动态目标空间位置的测量和自动跟踪等研究引起国内外厂商和学者的重视并进行了一系列探索性的研究。本课题主要研究复杂背景中对单个运动目标进行实时跟踪的方法。我们建立了一个基于系统，首先求解出基于梯度方法的光流场(原始图像中的运动区域)，然后根据诸如区域合并，区域连通或聚类处理等方法从光流场中抽取出运动目标区域，进而控制摄像头的转动使得运动目标始终位于摄像机的视野中，从而完成对复杂背景下单个运动目标的实时跟踪。实验结果表明，该方法能够在复杂运动的背景中基本实现对快速移动的目标的实时跟踪，例如，能够可靠地从风吹摆动的树丛中追踪走动的人，或在草丛中跟踪快速驰过的玩具汽车等。更多还原

【Abstract】 With the development of computer technology, the image processing is applied more and more extensively in almost fields of national economy. Recently, many companies and scholars at home and abroad think highly of tracking the moving objects automatically in multiple scenes. This paper describes a method that extracts the optical flows from sequent images and tracks the moving objects in real time. We set up a DSPs based on the TMS3206701.First, the DSP boards calculate optical flows (the velocity field of an image) based on the gradient method. Next, the region of moving object on the image is extracted from the optical flows using knowledge about the region of the object, such as Region-Merge arithmetic, Region-Connection arithmetic or Clustering arithmetic. Finally, the camera head is controlled to keep the object in the field of view of the camera. Therefore we achieve a real time tracking system to the moving objects in the multiple scenes. The results of experiments show that the method can track the moving objects in the multiple backgrounds in real time. For example, the walker walks through the avenue where the trees swayed in the wind, and the toy car speeds up across the brushwood.更多还原