In this thesis, we implement a face contour tracking system using Snakes Model. Snakes Model is an active contour model defined by an energy function in terms of the internal energy, the image energy, and the external energy. Then we can find the face contour model in the image by minimizing the energy function using dynamic programming.
We modified the energy function of Snakes model for the face contour tracking system. The continuities of the motion and the brightness are added to the energy function. And we proposed the new internal energy controlled by image energy, which makes the tracking system find the contour more exactly.
EKF(Extended Kalman Filter) is the general tool to estimate the object motion. We estimate the face contour model on the next image using EKF. It lessens the failure of the tracking system using Snakes model.
The proposed method can be applied to the face tracking in the virtual holography and the teleconferencing. And the extension of the proposed method helps to track other objects in the image.
