背景下运目跟踪算法的研究-沈阳航空航天大学毕业设计(论文).docVIP

简单背景下运动目标跟踪算法的研究 学 院 自动化学院 专 业 班 级 学 号 姓 名 指导教师 负责教师 沈阳航空2011年6月摘 要 关键词：Detection and tracking of moving object is an analytical technology studying how to extract the region of moving object,analysis the feature and motion of moving target,then match the object and estimate the concerned moving parameter in serial video image.Its an active branches in the area of image processing and computer vision.This technology has been applied widely in many ways such as security supervisory of important location,automatic driving,control of traffic flowing,the guidance of aero and military aircraft.In the inspection of the target motion, first ,using the current frame minus the background to get image of difference, Binarize the image by setting threshold to extract moving targets,then obtained complete reliable image of target, at last,with the target image post-processing on morphological filtering to eliminate the impact of noise and background disturbance;In the aspect of tracking moving target ,this article proposed two methods centroid tracking algorithm and kalman filter tracking algorithm to tracking motion.Centroid algorithm through calculating center of mass and radius r of each images white object,using centroid as the centre of a circle , r as radius to make a round frame as tracking window catching moving target , so as to search out moving object from image sequence more reliably , realized centroid tracking; The second kind is kalman filtering prediction algorithm,first, determine the target coordinates of first frame image , using this predicting function forecast goals position in the next frame through calculating the recursive formula, compared with the original data,error is very small. In this article, various design scheme and algorithm had been simulated in experiment and realized programming ,with the final determined algorithm repeatedly optimizing, greatly improving stability and instantaneity of the system . Experiments show that