C.-T. Kim, T.-Y. Choi, and J.-J. Lee (Korea)
Active Contour Model, Dynamic Programming, Energy Minimization, Object Tracking
An active contour models for object tracking using prior contour are investigated.The proposed algorithm uses edge with color feature for tracking and also uses the prior con tour information for robust characteristics.The orthogonal search to contour extends search range and reduces com putational burden compared with conventional snake al gorithms. The conventional algorithms have many prob lem for tracking, so model based techniques are noticed in tracking with snake. The tracking with models such as eigenvector model and dynamic model are suggested. But they need complex procedures such as training, estimation and matching. The proposed algorithm suggests the track ing scheme such like Kass's one, energy minimization. But it uses the prior contour as a model defining smooth sim ilarities with two contours. The algorithm searches nor mally and uses dynamic programming to solve energy min imization problem. The main drawbacks of conventional snake algorithm - shrinking, limited search range, sensi tive to noises - improved with the proposed algorithm. The external energy using color is also introduced. It can be adopted in robocup to detect ball or patch-tracking applica tion because of the ability to be able to use shape and color information in the same time
Important Links:
Go Back
