Dissertation
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 on SNAKE Model for Object Contour Extracting and Tracking

Author JinXueJun
Tutor LiWei;CaiJiaZuo;ZhengHeRong
School Zhejiang University of Technology
Course Applied Computer Technology
Keywords active contour module ghough transform dynamic programming kalman filter
CLC TP391.41
Type Master's thesis
Year 2006
Downloads 489
Quotes 0
Download Dissertation

The technology of Object Contour extraction is widely used in Image segmentation particularly medical imaging community , Motion tracking, Robot trace Planning, Monitoring System, No Driver Car, Cartoon Making and so on. Snake Model as an efficient way of implementing Object Contour extraction, is most used in follow applications:Image segmentation particularly medical imaging community , Motion tracking,Shape recognition.This dissertation presents an approach in tracking moving objects in a video sequence,named active contour module also named snake module,and also presents other improved approaches based on snake,and then integrated with Kalman Filter perform motion prediction, implementing tracking moving object.First,user initialise a contour close to the object of interest,following which,the contour should evolve and lock on to the object.Once a lock is obtained,its motion should be tracked over subsequent frames without any further user input necessary. This dissertation focus on object contour extraction. In this dissertation,based on the systematically analysis of theory and application research on the models,some improved approaches is proposed,such as using MINIMAX criterion to solve the problem of automatic selecting the parameters of the model,and using impvoed HOUGH transform to solve the problem of selecting initial contour,using impvoed dynamic programming in order to perform better convergence,and also using Kalman Filter to predict the possible position of moving target. The main works in this dissertation are as follows: 1. Implemeting greedy algorithm for snake.2. Solving the problem of selecting initial contour, solving the problem of automatic selecting the parameters of the model.3.Improving snake model, to makesnake model perform better convergenceAUsing kalman Filter to predict the possible position of moving target, Snake model integrate with Kalman filter to track moving object in a video sequence.

Related Dissertations
More Dissertations