Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

Research of Moving Object Detection and Tracking Technology

Author LiZeFeng
Tutor ZhuLingYun
School Donghua University
Course Control Theory and Control Engineering
Keywords Moving target detection Moving Target Tracking Gaussian mixture model Kalman filtering
CLC TP391.41
Type Master's thesis
Year 2012
Downloads 60
Quotes 0
Download Dissertation

Moving target detection and moving object tracking is one of the hot issues in the field of computer vision research , is also one of the key technologies , it combines image processing , pattern recognition , artificial intelligence as well as automatic control , and many other areas of knowledge , is an interdisciplinary the challenging topics at the forefront , has a wide range of applications in the field of intelligent visual surveillance , intelligent transportation , human-computer interaction , video compression , medical image analysis , robot vision navigation and other , therefore , the subject of study has important theoretical significance practical value and broad prospects for development . Moving target detection and tracking algorithm is the focus of this study , the static background target detection and tracking technology , such as video surveillance systems , traffic control systems , research has matured in the military, transportation , industrial manufacturing , medical , and everyday life of many areas has been widely used on the basis of the results of previous studies , the improvement and implementation of the moving target detection and tracking algorithms , mainly to complete the following tasks : ( 1 ) In-depth analysis of the movement of several common principles and characteristics of the target detection algorithm ; (2 ) an improved Gaussian mixture background modeling algorithm , using a constant learning rate variance Gaussian mixture model , over time , into a kind of variance a vicious cycle , the variance has been small ; ( 3) on several commonly used moving target detection algorithm research , through a large number of experiments , the effect of detection for various algorithms , real-time , the algorithm complexity , and complexity summarized the environmental adaptability performance analysis and comparison ; (4) focus on a moving object tracking based on Kalman filtering , gives a moving target tracking algorithm based on Kalman filtering , target detection , target motion estimation and object matching and tracking of three modules moving target tracking , accurate and stable tracking of moving targets .

Related Dissertations
More Dissertations